agora inbox for [email protected]help / color / mirror / Atom feed
[PATCH 1/6] sequential scan for dshash 63+ messages / 3 participants [nested] [flat]
* [PATCH 1/6] sequential scan for dshash @ 2018-06-29 07:41 Kyotaro Horiguchi <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Kyotaro Horiguchi @ 2018-06-29 07:41 UTC (permalink / raw) Add sequential scan feature to dshash. --- src/backend/lib/dshash.c | 188 ++++++++++++++++++++++++++++++++++++++++++++++- src/include/lib/dshash.h | 23 +++++- 2 files changed, 206 insertions(+), 5 deletions(-) diff --git a/src/backend/lib/dshash.c b/src/backend/lib/dshash.c index f095196fb6..1e8c22f94f 100644 --- a/src/backend/lib/dshash.c +++ b/src/backend/lib/dshash.c @@ -112,6 +112,7 @@ struct dshash_table size_t size_log2; /* log2(number of buckets) */ bool find_locked; /* Is any partition lock held by 'find'? */ bool find_exclusively_locked; /* ... exclusively? */ + bool seqscan_running;/* now under sequential scan */ }; /* Given a pointer to an item, find the entry (user data) it holds. */ @@ -127,6 +128,10 @@ struct dshash_table #define NUM_SPLITS(size_log2) \ (size_log2 - DSHASH_NUM_PARTITIONS_LOG2) +/* How many buckets are there in a given size? */ +#define NUM_BUCKETS(size_log2) \ + (((size_t) 1) << (size_log2)) + /* How many buckets are there in each partition at a given size? */ #define BUCKETS_PER_PARTITION(size_log2) \ (((size_t) 1) << NUM_SPLITS(size_log2)) @@ -153,6 +158,10 @@ struct dshash_table #define BUCKET_INDEX_FOR_PARTITION(partition, size_log2) \ ((partition) << NUM_SPLITS(size_log2)) +/* Choose partition based on bucket index. */ +#define PARTITION_FOR_BUCKET_INDEX(bucket_idx, size_log2) \ + ((bucket_idx) >> NUM_SPLITS(size_log2)) + /* The head of the active bucket for a given hash value (lvalue). */ #define BUCKET_FOR_HASH(hash_table, hash) \ (hash_table->buckets[ \ @@ -228,6 +237,7 @@ dshash_create(dsa_area *area, const dshash_parameters *params, void *arg) hash_table->find_locked = false; hash_table->find_exclusively_locked = false; + hash_table->seqscan_running = false; /* * Set up the initial array of buckets. Our initial size is the same as @@ -279,6 +289,7 @@ dshash_attach(dsa_area *area, const dshash_parameters *params, hash_table->control = dsa_get_address(area, control); hash_table->find_locked = false; hash_table->find_exclusively_locked = false; + hash_table->seqscan_running = false; Assert(hash_table->control->magic == DSHASH_MAGIC); /* @@ -324,7 +335,7 @@ dshash_destroy(dshash_table *hash_table) ensure_valid_bucket_pointers(hash_table); /* Free all the entries. */ - size = ((size_t) 1) << hash_table->size_log2; + size = NUM_BUCKETS(hash_table->size_log2); for (i = 0; i < size; ++i) { dsa_pointer item_pointer = hash_table->buckets[i]; @@ -549,9 +560,14 @@ dshash_delete_entry(dshash_table *hash_table, void *entry) LW_EXCLUSIVE)); delete_item(hash_table, item); - hash_table->find_locked = false; - hash_table->find_exclusively_locked = false; - LWLockRelease(PARTITION_LOCK(hash_table, partition)); + + /* We need to keep partition lock while sequential scan */ + if (!hash_table->seqscan_running) + { + hash_table->find_locked = false; + hash_table->find_exclusively_locked = false; + LWLockRelease(PARTITION_LOCK(hash_table, partition)); + } } /* @@ -568,6 +584,8 @@ dshash_release_lock(dshash_table *hash_table, void *entry) Assert(LWLockHeldByMeInMode(PARTITION_LOCK(hash_table, partition_index), hash_table->find_exclusively_locked ? LW_EXCLUSIVE : LW_SHARED)); + /* lock is under control of sequential scan */ + Assert(!hash_table->seqscan_running); hash_table->find_locked = false; hash_table->find_exclusively_locked = false; @@ -592,6 +610,168 @@ dshash_memhash(const void *v, size_t size, void *arg) return tag_hash(v, size); } +/* + * dshash_seq_init/_next/_term + * Sequentially scan trhough dshash table and return all the + * elements one by one, return NULL when no more. + * + * dshash_seq_term should be called if and only if the scan is abandoned + * before completion; if dshash_seq_next returns NULL then it has already done + * the end-of-scan cleanup. + * + * On returning element, it is locked as is the case with dshash_find. + * However, the caller must not release the lock. The lock is released as + * necessary in continued scan. + * + * As opposed to the equivalent for dynanash, the caller is not supposed to + * delete the returned element before continuing the scan. + * + * If consistent is set for dshash_seq_init, the whole hash table is + * non-exclusively locked. Otherwise a part of the hash table is locked in the + * same mode (partition lock). + */ +void +dshash_seq_init(dshash_seq_status *status, dshash_table *hash_table, + bool consistent, bool exclusive) +{ + /* allowed at most one scan at once */ + Assert(!hash_table->seqscan_running); + + status->hash_table = hash_table; + status->curbucket = 0; + status->nbuckets = 0; + status->curitem = NULL; + status->pnextitem = InvalidDsaPointer; + status->curpartition = -1; + status->consistent = consistent; + status->exclusive = exclusive; + hash_table->seqscan_running = true; + + /* + * Protect all partitions from modification if the caller wants a + * consistent result. + */ + if (consistent) + { + int i; + + for (i = 0; i < DSHASH_NUM_PARTITIONS; ++i) + { + Assert(!LWLockHeldByMe(PARTITION_LOCK(hash_table, i))); + + LWLockAcquire(PARTITION_LOCK(hash_table, i), + exclusive ? LW_EXCLUSIVE : LW_SHARED); + } + ensure_valid_bucket_pointers(hash_table); + } +} + +void * +dshash_seq_next(dshash_seq_status *status) +{ + dsa_pointer next_item_pointer; + + Assert(status->hash_table->seqscan_running); + if (status->curitem == NULL) + { + int partition; + + Assert (status->curbucket == 0); + Assert(!status->hash_table->find_locked); + + /* first shot. grab the first item. */ + if (!status->consistent) + { + partition = + PARTITION_FOR_BUCKET_INDEX(status->curbucket, + status->hash_table->size_log2); + LWLockAcquire(PARTITION_LOCK(status->hash_table, partition), + status->exclusive ? LW_EXCLUSIVE : LW_SHARED); + status->curpartition = partition; + + /* resize doesn't happen from now until seq scan ends */ + status->nbuckets = + NUM_BUCKETS(status->hash_table->control->size_log2); + ensure_valid_bucket_pointers(status->hash_table); + } + + next_item_pointer = status->hash_table->buckets[status->curbucket]; + } + else + next_item_pointer = status->pnextitem; + + /* Move to the next bucket if we finished the current bucket */ + while (!DsaPointerIsValid(next_item_pointer)) + { + if (++status->curbucket >= status->nbuckets) + { + /* all buckets have been scanned. finsih. */ + dshash_seq_term(status); + return NULL; + } + + /* Also move parititon lock if needed */ + if (!status->consistent) + { + int next_partition = + PARTITION_FOR_BUCKET_INDEX(status->curbucket, + status->hash_table->size_log2); + + /* Move lock along with partition for the bucket */ + if (status->curpartition != next_partition) + { + /* + * Take lock on the next partition then release the current, + * not in the reverse order. This is required to avoid + * resizing from happening during a sequential scan. Locks are + * taken in partition order so no dead lock happen with other + * seq scans or resizing. + */ + LWLockAcquire(PARTITION_LOCK(status->hash_table, + next_partition), + status->exclusive ? LW_EXCLUSIVE : LW_SHARED); + LWLockRelease(PARTITION_LOCK(status->hash_table, + status->curpartition)); + status->curpartition = next_partition; + } + } + + next_item_pointer = status->hash_table->buckets[status->curbucket]; + } + + status->curitem = + dsa_get_address(status->hash_table->area, next_item_pointer); + status->hash_table->find_locked = true; + status->hash_table->find_exclusively_locked = status->exclusive; + + /* + * This item can be deleted by the caller. Store the next item for the + * next iteration for the occasion. + */ + status->pnextitem = status->curitem->next; + + return ENTRY_FROM_ITEM(status->curitem); +} + +void +dshash_seq_term(dshash_seq_status *status) +{ + Assert(status->hash_table->seqscan_running); + status->hash_table->find_locked = false; + status->hash_table->find_exclusively_locked = false; + status->hash_table->seqscan_running = false; + + if (status->consistent) + { + int i; + + for (i = 0; i < DSHASH_NUM_PARTITIONS; ++i) + LWLockRelease(PARTITION_LOCK(status->hash_table, i)); + } + else if (status->curpartition >= 0) + LWLockRelease(PARTITION_LOCK(status->hash_table, status->curpartition)); +} + /* * Print debugging information about the internal state of the hash table to * stderr. The caller must hold no partition locks. diff --git a/src/include/lib/dshash.h b/src/include/lib/dshash.h index e5dfd57f0a..b80f3af995 100644 --- a/src/include/lib/dshash.h +++ b/src/include/lib/dshash.h @@ -59,6 +59,23 @@ typedef struct dshash_parameters struct dshash_table_item; typedef struct dshash_table_item dshash_table_item; +/* + * Sequential scan state of dshash. The detail is exposed since the storage + * size should be known to users but it should be considered as an opaque + * type by callers. + */ +typedef struct dshash_seq_status +{ + dshash_table *hash_table; + int curbucket; + int nbuckets; + dshash_table_item *curitem; + dsa_pointer pnextitem; + int curpartition; + bool consistent; + bool exclusive; +} dshash_seq_status; + /* Creating, sharing and destroying from hash tables. */ extern dshash_table *dshash_create(dsa_area *area, const dshash_parameters *params, @@ -70,7 +87,6 @@ extern dshash_table *dshash_attach(dsa_area *area, extern void dshash_detach(dshash_table *hash_table); extern dshash_table_handle dshash_get_hash_table_handle(dshash_table *hash_table); extern void dshash_destroy(dshash_table *hash_table); - /* Finding, creating, deleting entries. */ extern void *dshash_find(dshash_table *hash_table, const void *key, bool exclusive); @@ -80,6 +96,11 @@ extern bool dshash_delete_key(dshash_table *hash_table, const void *key); extern void dshash_delete_entry(dshash_table *hash_table, void *entry); extern void dshash_release_lock(dshash_table *hash_table, void *entry); +/* seq scan support */ +extern void dshash_seq_init(dshash_seq_status *status, dshash_table *hash_table, + bool consistent, bool exclusive); +extern void *dshash_seq_next(dshash_seq_status *status); +extern void dshash_seq_term(dshash_seq_status *status); /* Convenience hash and compare functions wrapping memcmp and tag_hash. */ extern int dshash_memcmp(const void *a, const void *b, size_t size, void *arg); extern dshash_hash dshash_memhash(const void *v, size_t size, void *arg); -- 2.16.3 ----Next_Part(Thu_Feb_21_16_05_55_2019_560)-- Content-Type: Text/X-Patch; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="v16-0002-Add-conditional-lock-feature-to-dshash.patch" ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH v1 6/7] Row pattern recognition patch (tests). @ 2023-06-25 11:48 Tatsuo Ishii <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Tatsuo Ishii @ 2023-06-25 11:48 UTC (permalink / raw) --- src/test/regress/expected/rpr.out | 273 +++++++++++++++++++++++++++++ src/test/regress/parallel_schedule | 2 +- src/test/regress/sql/rpr.sql | 150 ++++++++++++++++ 3 files changed, 424 insertions(+), 1 deletion(-) create mode 100644 src/test/regress/expected/rpr.out create mode 100644 src/test/regress/sql/rpr.sql diff --git a/src/test/regress/expected/rpr.out b/src/test/regress/expected/rpr.out new file mode 100644 index 0000000000..9ede60ba39 --- /dev/null +++ b/src/test/regress/expected/rpr.out @@ -0,0 +1,273 @@ +-- +-- Test for row pattern definition clause +-- +CREATE TEMP TABLE stock ( + company TEXT, + tdate DATE, + price INTEGER + ); +INSERT INTO stock VALUES ('company1', '2023-07-01', 100); +INSERT INTO stock VALUES ('company1', '2023-07-02', 200); +INSERT INTO stock VALUES ('company1', '2023-07-03', 150); +INSERT INTO stock VALUES ('company1', '2023-07-04', 140); +INSERT INTO stock VALUES ('company1', '2023-07-05', 150); +INSERT INTO stock VALUES ('company1', '2023-07-06', 90); +INSERT INTO stock VALUES ('company1', '2023-07-07', 110); +INSERT INTO stock VALUES ('company1', '2023-07-08', 130); +INSERT INTO stock VALUES ('company1', '2023-07-09', 120); +INSERT INTO stock VALUES ('company1', '2023-07-10', 130); +INSERT INTO stock VALUES ('company2', '2023-07-01', 50); +INSERT INTO stock VALUES ('company2', '2023-07-02', 2000); +INSERT INTO stock VALUES ('company2', '2023-07-03', 1500); +INSERT INTO stock VALUES ('company2', '2023-07-04', 1400); +INSERT INTO stock VALUES ('company2', '2023-07-05', 1500); +INSERT INTO stock VALUES ('company2', '2023-07-06', 60); +INSERT INTO stock VALUES ('company2', '2023-07-07', 1100); +INSERT INTO stock VALUES ('company2', '2023-07-08', 1300); +INSERT INTO stock VALUES ('company2', '2023-07-09', 1200); +INSERT INTO stock VALUES ('company2', '2023-07-10', 1300); +SELECT * FROM stock; + company | tdate | price +----------+------------+------- + company1 | 07-01-2023 | 100 + company1 | 07-02-2023 | 200 + company1 | 07-03-2023 | 150 + company1 | 07-04-2023 | 140 + company1 | 07-05-2023 | 150 + company1 | 07-06-2023 | 90 + company1 | 07-07-2023 | 110 + company1 | 07-08-2023 | 130 + company1 | 07-09-2023 | 120 + company1 | 07-10-2023 | 130 + company2 | 07-01-2023 | 50 + company2 | 07-02-2023 | 2000 + company2 | 07-03-2023 | 1500 + company2 | 07-04-2023 | 1400 + company2 | 07-05-2023 | 1500 + company2 | 07-06-2023 | 60 + company2 | 07-07-2023 | 1100 + company2 | 07-08-2023 | 1300 + company2 | 07-09-2023 | 1200 + company2 | 07-10-2023 | 1300 +(20 rows) + +-- basic test using PREV +SELECT company, tdate, price, rpr(price) OVER w FROM stock + WINDOW w AS ( + PARTITION BY company + ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING + INITIAL + PATTERN (START UP+ DOWN+) + DEFINE + START AS TRUE, + UP AS price > PREV(price), + DOWN AS price < PREV(price) +); + company | tdate | price | rpr +----------+------------+-------+------ + company1 | 07-01-2023 | 100 | 100 + company1 | 07-02-2023 | 200 | + company1 | 07-03-2023 | 150 | + company1 | 07-04-2023 | 140 | 140 + company1 | 07-05-2023 | 150 | + company1 | 07-06-2023 | 90 | 90 + company1 | 07-07-2023 | 110 | 110 + company1 | 07-08-2023 | 130 | + company1 | 07-09-2023 | 120 | + company1 | 07-10-2023 | 130 | + company2 | 07-01-2023 | 50 | 50 + company2 | 07-02-2023 | 2000 | + company2 | 07-03-2023 | 1500 | + company2 | 07-04-2023 | 1400 | 1400 + company2 | 07-05-2023 | 1500 | + company2 | 07-06-2023 | 60 | 60 + company2 | 07-07-2023 | 1100 | 1100 + company2 | 07-08-2023 | 1300 | + company2 | 07-09-2023 | 1200 | + company2 | 07-10-2023 | 1300 | +(20 rows) + +-- the first row start with less than or equal to 100 +SELECT company, tdate, price, rpr(price) OVER w FROM stock + WINDOW w AS ( + PARTITION BY company + ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING + INITIAL + PATTERN (LOWPRICE UP+ DOWN+) + DEFINE + LOWPRICE AS price <= 100, + UP AS price > PREV(price), + DOWN AS price < PREV(price) +); + company | tdate | price | rpr +----------+------------+-------+----- + company1 | 07-01-2023 | 100 | 100 + company1 | 07-02-2023 | 200 | + company1 | 07-03-2023 | 150 | + company1 | 07-04-2023 | 140 | + company1 | 07-05-2023 | 150 | + company1 | 07-06-2023 | 90 | 90 + company1 | 07-07-2023 | 110 | + company1 | 07-08-2023 | 130 | + company1 | 07-09-2023 | 120 | + company1 | 07-10-2023 | 130 | + company2 | 07-01-2023 | 50 | 50 + company2 | 07-02-2023 | 2000 | + company2 | 07-03-2023 | 1500 | + company2 | 07-04-2023 | 1400 | + company2 | 07-05-2023 | 1500 | + company2 | 07-06-2023 | 60 | 60 + company2 | 07-07-2023 | 1100 | + company2 | 07-08-2023 | 1300 | + company2 | 07-09-2023 | 1200 | + company2 | 07-10-2023 | 1300 | +(20 rows) + +-- second row raises 120% +SELECT company, tdate, price, rpr(price) OVER w FROM stock + WINDOW w AS ( + PARTITION BY company + ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING + INITIAL + PATTERN (LOWPRICE UP+ DOWN+) + DEFINE + LOWPRICE AS price <= 100, + UP AS price > PREV(price) * 1.2, + DOWN AS price < PREV(price) +); + company | tdate | price | rpr +----------+------------+-------+----- + company1 | 07-01-2023 | 100 | 100 + company1 | 07-02-2023 | 200 | + company1 | 07-03-2023 | 150 | + company1 | 07-04-2023 | 140 | + company1 | 07-05-2023 | 150 | + company1 | 07-06-2023 | 90 | + company1 | 07-07-2023 | 110 | + company1 | 07-08-2023 | 130 | + company1 | 07-09-2023 | 120 | + company1 | 07-10-2023 | 130 | + company2 | 07-01-2023 | 50 | 50 + company2 | 07-02-2023 | 2000 | + company2 | 07-03-2023 | 1500 | + company2 | 07-04-2023 | 1400 | + company2 | 07-05-2023 | 1500 | + company2 | 07-06-2023 | 60 | + company2 | 07-07-2023 | 1100 | + company2 | 07-08-2023 | 1300 | + company2 | 07-09-2023 | 1200 | + company2 | 07-10-2023 | 1300 | +(20 rows) + +-- using NEXT +SELECT company, tdate, price, rpr(price) OVER w FROM stock + WINDOW w AS ( + PARTITION BY company + ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING + INITIAL + PATTERN (START UPDOWN) + DEFINE + START AS TRUE, + UPDOWN AS price > PREV(price) AND price > NEXT(price) +); + company | tdate | price | rpr +----------+------------+-------+------ + company1 | 07-01-2023 | 100 | 100 + company1 | 07-02-2023 | 200 | + company1 | 07-03-2023 | 150 | + company1 | 07-04-2023 | 140 | 140 + company1 | 07-05-2023 | 150 | + company1 | 07-06-2023 | 90 | + company1 | 07-07-2023 | 110 | 110 + company1 | 07-08-2023 | 130 | + company1 | 07-09-2023 | 120 | + company1 | 07-10-2023 | 130 | + company2 | 07-01-2023 | 50 | 50 + company2 | 07-02-2023 | 2000 | + company2 | 07-03-2023 | 1500 | + company2 | 07-04-2023 | 1400 | 1400 + company2 | 07-05-2023 | 1500 | + company2 | 07-06-2023 | 60 | + company2 | 07-07-2023 | 1100 | 1100 + company2 | 07-08-2023 | 1300 | + company2 | 07-09-2023 | 1200 | + company2 | 07-10-2023 | 1300 | +(20 rows) + +-- row pattern definition variable name must not appear more than once +SELECT company, tdate, price, rpr(price) OVER w FROM stock + WINDOW w AS ( + PARTITION BY company + ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING + INITIAL + PATTERN (START UP+ DOWN+) + DEFINE + START AS TRUE, + UP AS price > PREV(price), + DOWN AS price < PREV(price), + UP AS price > PREV(price) +); +ERROR: row pattern definition variable name "up" appears more than once in DEFINE clause +LINE 9: UP AS price > PREV(price), + ^ +-- pattern variable name must appear in DEFINE +SELECT company, tdate, price, rpr(price) OVER w FROM stock + WINDOW w AS ( + PARTITION BY company + ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING + INITIAL + PATTERN (START UP+ DOWN+ END) + DEFINE + START AS TRUE, + UP AS price > PREV(price), + DOWN AS price < PREV(price) +); +ERROR: syntax error at or near "END" +LINE 6: PATTERN (START UP+ DOWN+ END) + ^ +-- FRAME must start at current row when row patttern recognition is used +SELECT company, tdate, price, rpr(price) OVER w FROM stock + WINDOW w AS ( + PARTITION BY company + ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING + INITIAL + PATTERN (START UP+ DOWN+) + DEFINE + START AS TRUE, + UP AS price > PREV(price), + DOWN AS price < PREV(price) +); +ERROR: FRAME must start at current row when row patttern recognition is used +-- AFTER MATCH SKIP TO PAST LAST ROW is not supported +SELECT company, tdate, price, rpr(price) OVER w FROM stock + WINDOW w AS ( + PARTITION BY company + ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING + AFTER MATCH SKIP TO PAST LAST ROW + INITIAL + PATTERN (START UP+ DOWN+) + DEFINE + START AS TRUE, + UP AS price > PREV(price), + DOWN AS price < PREV(price) +); +ERROR: syntax error at or near "PAST" +LINE 5: AFTER MATCH SKIP TO PAST LAST ROW + ^ +-- SEEK is not supported +SELECT company, tdate, price, rpr(price) OVER w FROM stock + WINDOW w AS ( + PARTITION BY company + ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING + AFTER MATCH SKIP TO NEXT ROW + SEEK + PATTERN (START UP+ DOWN+) + DEFINE + START AS TRUE, + UP AS price > PREV(price), + DOWN AS price < PREV(price) +); +ERROR: SEEK is not supported +LINE 6: SEEK + ^ +HINT: Use INITIAL. diff --git a/src/test/regress/parallel_schedule b/src/test/regress/parallel_schedule index cf46fa3359..ebb741318a 100644 --- a/src/test/regress/parallel_schedule +++ b/src/test/regress/parallel_schedule @@ -98,7 +98,7 @@ test: publication subscription # Another group of parallel tests # select_views depends on create_view # ---------- -test: select_views portals_p2 foreign_key cluster dependency guc bitmapops combocid tsearch tsdicts foreign_data window xmlmap functional_deps advisory_lock indirect_toast equivclass +test: select_views portals_p2 foreign_key cluster dependency guc bitmapops combocid tsearch tsdicts foreign_data window xmlmap functional_deps advisory_lock indirect_toast equivclass rpr # ---------- # Another group of parallel tests (JSON related) diff --git a/src/test/regress/sql/rpr.sql b/src/test/regress/sql/rpr.sql new file mode 100644 index 0000000000..921a9fcdfa --- /dev/null +++ b/src/test/regress/sql/rpr.sql @@ -0,0 +1,150 @@ +-- +-- Test for row pattern definition clause +-- + +CREATE TEMP TABLE stock ( + company TEXT, + tdate DATE, + price INTEGER + ); +INSERT INTO stock VALUES ('company1', '2023-07-01', 100); +INSERT INTO stock VALUES ('company1', '2023-07-02', 200); +INSERT INTO stock VALUES ('company1', '2023-07-03', 150); +INSERT INTO stock VALUES ('company1', '2023-07-04', 140); +INSERT INTO stock VALUES ('company1', '2023-07-05', 150); +INSERT INTO stock VALUES ('company1', '2023-07-06', 90); +INSERT INTO stock VALUES ('company1', '2023-07-07', 110); +INSERT INTO stock VALUES ('company1', '2023-07-08', 130); +INSERT INTO stock VALUES ('company1', '2023-07-09', 120); +INSERT INTO stock VALUES ('company1', '2023-07-10', 130); +INSERT INTO stock VALUES ('company2', '2023-07-01', 50); +INSERT INTO stock VALUES ('company2', '2023-07-02', 2000); +INSERT INTO stock VALUES ('company2', '2023-07-03', 1500); +INSERT INTO stock VALUES ('company2', '2023-07-04', 1400); +INSERT INTO stock VALUES ('company2', '2023-07-05', 1500); +INSERT INTO stock VALUES ('company2', '2023-07-06', 60); +INSERT INTO stock VALUES ('company2', '2023-07-07', 1100); +INSERT INTO stock VALUES ('company2', '2023-07-08', 1300); +INSERT INTO stock VALUES ('company2', '2023-07-09', 1200); +INSERT INTO stock VALUES ('company2', '2023-07-10', 1300); + +SELECT * FROM stock; + +-- basic test using PREV +SELECT company, tdate, price, rpr(price) OVER w FROM stock + WINDOW w AS ( + PARTITION BY company + ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING + INITIAL + PATTERN (START UP+ DOWN+) + DEFINE + START AS TRUE, + UP AS price > PREV(price), + DOWN AS price < PREV(price) +); + +-- the first row start with less than or equal to 100 +SELECT company, tdate, price, rpr(price) OVER w FROM stock + WINDOW w AS ( + PARTITION BY company + ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING + INITIAL + PATTERN (LOWPRICE UP+ DOWN+) + DEFINE + LOWPRICE AS price <= 100, + UP AS price > PREV(price), + DOWN AS price < PREV(price) +); + +-- second row raises 120% +SELECT company, tdate, price, rpr(price) OVER w FROM stock + WINDOW w AS ( + PARTITION BY company + ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING + INITIAL + PATTERN (LOWPRICE UP+ DOWN+) + DEFINE + LOWPRICE AS price <= 100, + UP AS price > PREV(price) * 1.2, + DOWN AS price < PREV(price) +); + +-- using NEXT +SELECT company, tdate, price, rpr(price) OVER w FROM stock + WINDOW w AS ( + PARTITION BY company + ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING + INITIAL + PATTERN (START UPDOWN) + DEFINE + START AS TRUE, + UPDOWN AS price > PREV(price) AND price > NEXT(price) +); + +-- row pattern definition variable name must not appear more than once +SELECT company, tdate, price, rpr(price) OVER w FROM stock + WINDOW w AS ( + PARTITION BY company + ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING + INITIAL + PATTERN (START UP+ DOWN+) + DEFINE + START AS TRUE, + UP AS price > PREV(price), + DOWN AS price < PREV(price), + UP AS price > PREV(price) +); + +-- pattern variable name must appear in DEFINE +SELECT company, tdate, price, rpr(price) OVER w FROM stock + WINDOW w AS ( + PARTITION BY company + ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING + INITIAL + PATTERN (START UP+ DOWN+ END) + DEFINE + START AS TRUE, + UP AS price > PREV(price), + DOWN AS price < PREV(price) +); + +-- FRAME must start at current row when row patttern recognition is used +SELECT company, tdate, price, rpr(price) OVER w FROM stock + WINDOW w AS ( + PARTITION BY company + ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING + INITIAL + PATTERN (START UP+ DOWN+) + DEFINE + START AS TRUE, + UP AS price > PREV(price), + DOWN AS price < PREV(price) +); + +-- AFTER MATCH SKIP TO PAST LAST ROW is not supported +SELECT company, tdate, price, rpr(price) OVER w FROM stock + WINDOW w AS ( + PARTITION BY company + ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING + AFTER MATCH SKIP TO PAST LAST ROW + INITIAL + PATTERN (START UP+ DOWN+) + DEFINE + START AS TRUE, + UP AS price > PREV(price), + DOWN AS price < PREV(price) +); + +-- SEEK is not supported +SELECT company, tdate, price, rpr(price) OVER w FROM stock + WINDOW w AS ( + PARTITION BY company + ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING + AFTER MATCH SKIP TO NEXT ROW + SEEK + PATTERN (START UP+ DOWN+) + DEFINE + START AS TRUE, + UP AS price > PREV(price), + DOWN AS price < PREV(price) +); -- 2.25.1 ----Next_Part(Sun_Jun_25_21_05_09_2023_126)---- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
* [PATCH 2/2] Publish list of tables being repacked in shared memory @ 2026-04-07 20:29 Álvaro Herrera <[email protected]> 0 siblings, 0 replies; 63+ messages in thread From: Álvaro Herrera @ 2026-04-07 20:29 UTC (permalink / raw) Use it in autovacuum to skip processing tables that are being repacked. This is mostly to avoid repeated attempts to process such tables, which would fail due to the special deadlock checker behavior for repack. Author: Álvaro Herrera <[email protected]> Discussion: https://postgr.es/m/[email protected] --- src/backend/commands/repack.c | 195 ++++++++++++++++-- src/backend/postmaster/autovacuum.c | 20 ++ .../utils/activity/wait_event_names.txt | 1 + src/include/commands/repack.h | 2 + src/include/storage/lwlocklist.h | 2 +- src/include/storage/subsystemlist.h | 1 + src/tools/pgindent/typedefs.list | 3 + 7 files changed, 210 insertions(+), 14 deletions(-) diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index a5f5df77291..ee7072dce6a 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -63,9 +63,11 @@ #include "optimizer/optimizer.h" #include "pgstat.h" #include "storage/bufmgr.h" +#include "storage/ipc.h" #include "storage/lmgr.h" #include "storage/predicate.h" #include "storage/proc.h" +#include "storage/subsystems.h" #include "utils/acl.h" #include "utils/fmgroids.h" #include "utils/guc.h" @@ -79,6 +81,32 @@ #include "utils/syscache.h" #include "utils/wait_event_types.h" + +/* Shared memory layout for REPACK */ +typedef struct RepackWorkerInfo +{ + bool ri_in_use; + pid_t ri_backendpid; + Oid ri_dbid; + Oid ri_relid; + Oid ri_toastrelid; +} RepackWorkerInfo; + +typedef struct +{ + bool re_useless; + RepackWorkerInfo re_workerinfo[FLEXIBLE_ARRAY_MEMBER]; +} RepackShmemStruct; + +static RepackShmemStruct *RepackShmem; + +typedef struct RepackCleanupContext +{ + bool concurrent; + int workerindex; +} RepackCleanupContext; + + /* * This struct is used to pass around the information on tables to be * clustered. We need this so we can make a list of them when invoked without @@ -90,6 +118,7 @@ typedef struct Oid indexOid; } RelToCluster; + /* * The first file exported by the decoding worker must contain a snapshot, the * following ones contain the data changes. @@ -166,6 +195,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd, MemoryContext permcxt); static bool repack_is_permitted_for_relation(RepackCommand cmd, Oid relid, Oid userid); +static void RepackCleanup(RepackCleanupContext *context); +static void RepackCleanupCb(int code, Datum arg); +static void RepackShmemRequest(void *arg); +static void RepackShmemInit(void *arg); static void apply_concurrent_changes(BufFile *file, ChangeContext *chgcxt); static void apply_concurrent_insert(Relation rel, TupleTableSlot *slot, @@ -210,6 +243,11 @@ static void ProcessRepackMessage(StringInfo msg); static const char *RepackCommandAsString(RepackCommand cmd); +const ShmemCallbacks RepackShmemCallbacks = { + .request_fn = RepackShmemRequest, + .init_fn = RepackShmemInit, +}; + /* * The repack code allows for processing multiple tables at once. Because * of this, we cannot just run everything on a single transaction, or we @@ -514,6 +552,7 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, Oid tableOid = RelationGetRelid(OldHeap); Relation index; LOCKMODE lmode; + RepackCleanupContext context; Oid save_userid; int save_sec_context; int save_nestlevel; @@ -660,24 +699,43 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid, TransferPredicateLocksToHeapRelation(OldHeap); /* rebuild_relation does all the dirty work */ - PG_TRY(); - { - rebuild_relation(OldHeap, index, verbose, ident_idx); - } - PG_FINALLY(); + context.concurrent = concurrent; + + PG_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); { if (concurrent) { - /* - * Since during normal operation the worker was already asked to - * exit, stopping it explicitly is especially important on ERROR. - * However it still seems a good practice to make sure that the - * worker never survives the REPACK command. - */ - stop_repack_decoding_worker(); + bool freefound = false; + + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *worker; + + if (RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + freefound = true; + worker = &RepackShmem->re_workerinfo[i]; + context.workerindex = i; + + worker->ri_in_use = true; + worker->ri_backendpid = MyProcPid; + worker->ri_dbid = MyDatabaseId; + worker->ri_relid = RelationGetRelid(OldHeap); + worker->ri_toastrelid = OldHeap->rd_rel->reltoastrelid; + break; + } + if (!freefound) + elog(ERROR, "could not find free repack entry"); + LWLockRelease(RepackLock); } + + rebuild_relation(OldHeap, index, verbose, ident_idx); } - PG_END_TRY(); + PG_END_ENSURE_ERROR_CLEANUP(RepackCleanupCb, PointerGetDatum(&context)); + + RepackCleanup(&context); /* rebuild_relation closes OldHeap, and index if valid */ @@ -691,6 +749,117 @@ out: pgstat_progress_end_command(); } +/* + * Return whether any backend is running concurrent REPACK on the given table + * (which could be a toast table). + */ +bool +is_table_under_repack(Oid databaseId, Oid relid) +{ + bool retval = false; + + LWLockAcquire(RepackLock, LW_SHARED); + for (int i = 0; i < max_repack_replication_slots; i++) + { + RepackWorkerInfo *rworker; + + if (!RepackShmem->re_workerinfo[i].ri_in_use) + continue; + + rworker = &RepackShmem->re_workerinfo[i]; + if (rworker->ri_dbid == MyDatabaseId && + (rworker->ri_relid == relid || + rworker->ri_toastrelid == relid)) + retval = true; + } + LWLockRelease(RepackLock); + + return retval; +} + +/* + * Remove ourselves from the workerinfo array. + */ +static void +RepackCleanup(RepackCleanupContext *context) +{ + if (context->concurrent) + { + RepackWorkerInfo *worker; + + /* + * The worker would normally terminate on its own when the work is + * done, but make sure we signal it just in case. + */ + stop_repack_decoding_worker(); + + /* + * also, make sure we stop advertising the relation we were repacking, + * so that autovacuum reverts to handling it normally. + */ + LWLockAcquire(RepackLock, LW_EXCLUSIVE); + + worker = &RepackShmem->re_workerinfo[context->workerindex]; + Assert(worker->ri_backendpid == MyProcPid); + worker->ri_in_use = false; + worker->ri_backendpid = 0; + worker->ri_dbid = InvalidOid; + worker->ri_relid = InvalidOid; + worker->ri_toastrelid = InvalidOid; + LWLockRelease(RepackLock); + } +} + +/* + * RepackCleanup wrapped as an on_shmem_exit callback function + */ +static void +RepackCleanupCb(int code, Datum arg) +{ + RepackCleanup((RepackCleanupContext *) DatumGetPointer(arg)); +} + +/* + * RepackShmemRequest + * Register shared memory space needed for repack + */ +static void +RepackShmemRequest(void *arg) +{ + Size size; + + /* + * Need the fixed struct and the array of RepackWorkerInfo. + */ + size = sizeof(RepackShmemStruct); + size = MAXALIGN(size); + size = add_size(size, mul_size(max_repack_replication_slots, + sizeof(RepackWorkerInfo))); + + ShmemRequestStruct(.name = "Repack Data", + .size = size, + .ptr = (void **) &RepackShmem, + ); +} + +static void +RepackShmemInit(void *arg) +{ + RepackWorkerInfo *reinfo; + + reinfo = (RepackWorkerInfo *) ((char *) RepackShmem + + MAXALIGN(sizeof(RepackShmemStruct))); + + for (int i = 0; i < max_repack_replication_slots; i++) + { + reinfo[i].ri_in_use = false; + reinfo[i].ri_backendpid = 0; + reinfo[i].ri_dbid = InvalidOid; + reinfo[i].ri_relid = InvalidOid; + reinfo[i].ri_toastrelid = InvalidOid; + } +} + /* * Check if the table (and its index) still meets the requirements of * cluster_rel(). diff --git a/src/backend/postmaster/autovacuum.c b/src/backend/postmaster/autovacuum.c index bd626a16363..080c64ea3c8 100644 --- a/src/backend/postmaster/autovacuum.c +++ b/src/backend/postmaster/autovacuum.c @@ -78,6 +78,7 @@ #include "catalog/namespace.h" #include "catalog/pg_database.h" #include "catalog/pg_namespace.h" +#include "commands/repack.h" #include "commands/vacuum.h" #include "common/int.h" #include "funcapi.h" @@ -2422,6 +2423,25 @@ do_autovacuum(void) } } LWLockRelease(AutovacuumLock); + + /* + * Similarly, if the table is being processed by concurrent repack, + * skip it (but make a note of that). We wouldn't be able to acquire + * its lock anyway. + */ + if (!skipit) + { + MemoryContextSwitchTo(PortalContext); + + skipit = is_table_under_repack(MyDatabaseId, relid); + if (skipit) + ereport(LOG, + errmsg("skipping table \"%s.%s.%s\" because it's being repacked in concurrent mode", + get_database_name(MyDatabaseId), + get_namespace_name(get_rel_namespace(relid)), + get_rel_name(relid))); + } + if (skipit) { LWLockRelease(AutovacuumScheduleLock); diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt index 7bda5298558..e206304f204 100644 --- a/src/backend/utils/activity/wait_event_names.txt +++ b/src/backend/utils/activity/wait_event_names.txt @@ -332,6 +332,7 @@ SInvalWrite "Waiting to add a message to the shared catalog invalidation queue." WALBufMapping "Waiting to replace a page in WAL buffers." WALWrite "Waiting for WAL buffers to be written to disk." ControlFile "Waiting to read or update the <filename>pg_control</filename> file or create a new WAL file." +Repack "Waiting to read or update tables in process by concurrent repack." MultiXactGen "Waiting to read or update shared multixact state." RelCacheInit "Waiting to read or update a <filename>pg_internal.init</filename> relation cache initialization file." CheckpointerComm "Waiting to manage fsync requests." diff --git a/src/include/commands/repack.h b/src/include/commands/repack.h index fd16e74b179..be7d38b5fae 100644 --- a/src/include/commands/repack.h +++ b/src/include/commands/repack.h @@ -42,6 +42,8 @@ extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel); extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid, ClusterParams *params, bool isTopLevel); +extern bool is_table_under_repack(Oid databaseId, Oid relid); + extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid, LOCKMODE lockmode); extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal); diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h index af8553bcb6c..3f08f4a15d4 100644 --- a/src/include/storage/lwlocklist.h +++ b/src/include/storage/lwlocklist.h @@ -41,7 +41,7 @@ PG_LWLOCK(6, SInvalWrite) PG_LWLOCK(7, WALBufMapping) PG_LWLOCK(8, WALWrite) PG_LWLOCK(9, ControlFile) -/* 10 was CheckpointLock */ +PG_LWLOCK(10, Repack) /* 11 was XactSLRULock */ /* 12 was SubtransSLRULock */ PG_LWLOCK(13, MultiXactGen) diff --git a/src/include/storage/subsystemlist.h b/src/include/storage/subsystemlist.h index 9ad619080be..4e683b8b0a8 100644 --- a/src/include/storage/subsystemlist.h +++ b/src/include/storage/subsystemlist.h @@ -72,6 +72,7 @@ PG_SHMEM_SUBSYSTEM(WalSummarizerShmemCallbacks) PG_SHMEM_SUBSYSTEM(PgArchShmemCallbacks) PG_SHMEM_SUBSYSTEM(ApplyLauncherShmemCallbacks) PG_SHMEM_SUBSYSTEM(SlotSyncShmemCallbacks) +PG_SHMEM_SUBSYSTEM(RepackShmemCallbacks) /* other modules that need some shared memory space */ PG_SHMEM_SUBSYSTEM(BTreeShmemCallbacks) diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index 637c669a146..d019e03aaf1 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -2639,9 +2639,12 @@ ReorderBufferTupleCidEnt ReorderBufferTupleCidKey ReorderBufferUpdateProgressTxnCB ReorderTuple +RepackCleanupContext RepackCommand RepackDecodingState +RepackShmemStruct RepackStmt +RepackWorkerInfo ReparameterizeForeignPathByChild_function ReplOriginId ReplOriginXactState -- 2.47.3 --kdrcpfmkbkc4lqhu-- ^ permalink raw reply [nested|flat] 63+ messages in thread
end of thread, other threads:[~2026-04-07 20:29 UTC | newest] Thread overview: 63+ messages (download: mbox mbox.gz follow: Atom feed) -- links below jump to the message on this page -- 2018-06-29 07:41 [PATCH 1/6] sequential scan for dshash Kyotaro Horiguchi <[email protected]> 2023-06-25 11:48 [PATCH v1 6/7] Row pattern recognition patch (tests). Tatsuo Ishii <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]> 2026-04-07 20:29 [PATCH 2/2] Publish list of tables being repacked in shared memory Álvaro Herrera <[email protected]>
This inbox is served by agora; see mirroring instructions for how to clone and mirror all data and code used for this inbox