agora inbox for pgsql-hackers@postgresql.orghelp / color / mirror / Atom feed
[PATCH v25 1/4] vacuum errcontext to show block being processed 10+ messages / 2 participants [nested] [flat]
* [PATCH v17 1/2] vacuum errcontext to show block being processed @ 2019-12-13 02:54 Justin Pryzby <pryzbyj@telsasoft.com> 0 siblings, 0 replies; 10+ messages in thread From: Justin Pryzby @ 2019-12-13 02:54 UTC (permalink / raw) As requested here. https://www.postgresql.org/message-id/20190807235154.erbmr4o4bo6vgnjv%40alap3.anarazel.de --- src/backend/access/heap/vacuumlazy.c | 117 +++++++++++++++++++++++++++++++++++ 1 file changed, 117 insertions(+) diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c index a23cdef..9358ab4 100644 --- a/src/backend/access/heap/vacuumlazy.c +++ b/src/backend/access/heap/vacuumlazy.c @@ -292,6 +292,14 @@ typedef struct LVRelStats bool lock_waiter_detected; } LVRelStats; +typedef struct +{ + char *relnamespace; + char *relname; + char *indname; /* undefined while not processing index */ + BlockNumber blkno; /* undefined while not processing heap */ + int phase; /* Reusing same enums as for progress reporting */ +} vacuum_error_callback_arg; /* A few variables that don't seem worth passing around as parameters */ static int elevel = -1; @@ -361,6 +369,7 @@ static void end_parallel_vacuum(Relation *Irel, IndexBulkDeleteResult **stats, LVParallelState *lps, int nindexes); static LVSharedIndStats *get_indstats(LVShared *lvshared, int n); static bool skip_parallel_vacuum_index(Relation indrel, LVShared *lvshared); +static void vacuum_error_callback(void *arg); /* @@ -724,6 +733,8 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, PROGRESS_VACUUM_MAX_DEAD_TUPLES }; int64 initprog_val[3]; + ErrorContextCallback errcallback; + vacuum_error_callback_arg errcbarg; pg_rusage_init(&ru0); @@ -870,6 +881,17 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, else skipping_blocks = false; + /* Setup error traceback support for ereport() */ + errcbarg.relnamespace = get_namespace_name(RelationGetNamespace(onerel)); + errcbarg.relname = relname; + errcbarg.blkno = InvalidBlockNumber; /* Not known yet */ + errcbarg.phase = PROGRESS_VACUUM_PHASE_SCAN_HEAP; + + errcallback.callback = vacuum_error_callback; + errcallback.arg = (void *) &errcbarg; + errcallback.previous = error_context_stack; + error_context_stack = &errcallback; + for (blkno = 0; blkno < nblocks; blkno++) { Buffer buf; @@ -891,6 +913,8 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, #define FORCE_CHECK_PAGE() \ (blkno == nblocks - 1 && should_attempt_truncation(params, vacrelstats)) + errcbarg.blkno = blkno; + pgstat_progress_update_param(PROGRESS_VACUUM_HEAP_BLKS_SCANNED, blkno); if (blkno == next_unskippable_block) @@ -987,6 +1011,9 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, vmbuffer = InvalidBuffer; } + /* Pop the error context stack */ + error_context_stack = errcallback.previous; + /* Work on all the indexes, then the heap */ lazy_vacuum_all_indexes(onerel, Irel, indstats, vacrelstats, lps, nindexes); @@ -1011,6 +1038,9 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, /* Report that we are once again scanning the heap */ pgstat_progress_update_param(PROGRESS_VACUUM_PHASE, PROGRESS_VACUUM_PHASE_SCAN_HEAP); + + /* Set the error context while continuing heap scan */ + error_context_stack = &errcallback; } /* @@ -1597,6 +1627,9 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, RecordPageWithFreeSpace(onerel, blkno, freespace); } + /* Pop the error context stack */ + error_context_stack = errcallback.previous; + /* report that everything is scanned and vacuumed */ pgstat_progress_update_param(PROGRESS_VACUUM_HEAP_BLKS_SCANNED, blkno); @@ -1772,11 +1805,26 @@ lazy_vacuum_heap(Relation onerel, LVRelStats *vacrelstats) int npages; PGRUsage ru0; Buffer vmbuffer = InvalidBuffer; + ErrorContextCallback errcallback; + vacuum_error_callback_arg errcbarg; /* Report that we are now vacuuming the heap */ pgstat_progress_update_param(PROGRESS_VACUUM_PHASE, PROGRESS_VACUUM_PHASE_VACUUM_HEAP); + /* + * Setup error traceback support for ereport() + */ + errcbarg.relnamespace = get_namespace_name(RelationGetNamespace(onerel)); + errcbarg.relname = RelationGetRelationName(onerel); + errcbarg.blkno = InvalidBlockNumber; /* Not known yet */ + errcbarg.phase = PROGRESS_VACUUM_PHASE_VACUUM_HEAP; + + errcallback.callback = vacuum_error_callback; + errcallback.arg = (void *) &errcbarg; + errcallback.previous = error_context_stack; + error_context_stack = &errcallback; + pg_rusage_init(&ru0); npages = 0; @@ -1791,6 +1839,7 @@ lazy_vacuum_heap(Relation onerel, LVRelStats *vacrelstats) vacuum_delay_point(); tblk = ItemPointerGetBlockNumber(&vacrelstats->dead_tuples->itemptrs[tupindex]); + errcbarg.blkno = tblk; buf = ReadBufferExtended(onerel, MAIN_FORKNUM, tblk, RBM_NORMAL, vac_strategy); if (!ConditionalLockBufferForCleanup(buf)) @@ -1811,6 +1860,9 @@ lazy_vacuum_heap(Relation onerel, LVRelStats *vacrelstats) npages++; } + /* Pop the error context stack */ + error_context_stack = errcallback.previous; + if (BufferIsValid(vmbuffer)) { ReleaseBuffer(vmbuffer); @@ -2318,6 +2370,8 @@ lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, IndexVacuumInfo ivinfo; const char *msg; PGRUsage ru0; + ErrorContextCallback errcallback; + vacuum_error_callback_arg errcbarg; pg_rusage_init(&ru0); @@ -2329,10 +2383,24 @@ lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, ivinfo.num_heap_tuples = reltuples; ivinfo.strategy = vac_strategy; + /* Setup error traceback support for ereport() */ + errcbarg.relnamespace = get_namespace_name(RelationGetNamespace(indrel)); + errcbarg.indname = RelationGetRelationName(indrel); + errcbarg.relname = get_rel_name(indrel->rd_index->indexrelid); + errcbarg.phase = PROGRESS_VACUUM_PHASE_VACUUM_INDEX; + + errcallback.callback = vacuum_error_callback; + errcallback.arg = (void *) &errcbarg; + errcallback.previous = error_context_stack; + error_context_stack = &errcallback; + /* Do bulk deletion */ *stats = index_bulk_delete(&ivinfo, *stats, lazy_tid_reaped, (void *) dead_tuples); + /* Pop the error context stack */ + error_context_stack = errcallback.previous; + if (IsParallelWorker()) msg = gettext_noop("scanned index \"%s\" to remove %d row versions by parallel vacuum worker"); else @@ -2359,6 +2427,8 @@ lazy_cleanup_index(Relation indrel, IndexVacuumInfo ivinfo; const char *msg; PGRUsage ru0; + vacuum_error_callback_arg errcbarg; + ErrorContextCallback errcallback; pg_rusage_init(&ru0); @@ -2371,8 +2441,22 @@ lazy_cleanup_index(Relation indrel, ivinfo.num_heap_tuples = reltuples; ivinfo.strategy = vac_strategy; + /* Setup error traceback support for ereport() */ + errcbarg.relnamespace = get_namespace_name(RelationGetNamespace(indrel)); + errcbarg.indname = RelationGetRelationName(indrel); + errcbarg.relname = get_rel_name(indrel->rd_index->indexrelid); + errcbarg.phase = PROGRESS_VACUUM_PHASE_INDEX_CLEANUP; + + errcallback.previous = error_context_stack; + errcallback.callback = vacuum_error_callback; + errcallback.arg = (void *) &errcbarg; + error_context_stack = &errcallback; + *stats = index_vacuum_cleanup(&ivinfo, *stats); + /* Pop the error context stack */ + error_context_stack = errcallback.previous; + if (!(*stats)) return; @@ -3375,3 +3459,36 @@ parallel_vacuum_main(dsm_segment *seg, shm_toc *toc) table_close(onerel, ShareUpdateExclusiveLock); pfree(stats); } + +/* + * Error context callback for errors occurring during vacuum. + */ +static void +vacuum_error_callback(void *arg) +{ + vacuum_error_callback_arg *cbarg = arg; + + switch (cbarg->phase) { + case PROGRESS_VACUUM_PHASE_SCAN_HEAP: + if (BlockNumberIsValid(cbarg->blkno)) + errcontext(_("while scanning block %u of relation \"%s.%s\""), + cbarg->blkno, cbarg->relnamespace, cbarg->relname); + break; + + case PROGRESS_VACUUM_PHASE_VACUUM_HEAP: + if (BlockNumberIsValid(cbarg->blkno)) + errcontext(_("while vacuuming block %u of relation \"%s.%s\""), + cbarg->blkno, cbarg->relnamespace, cbarg->relname); + break; + + case PROGRESS_VACUUM_PHASE_VACUUM_INDEX: + errcontext(_("while vacuuming index \"%s.%s\" of relation \"%s\""), + cbarg->relnamespace, cbarg->indname, cbarg->relname); + break; + + case PROGRESS_VACUUM_PHASE_INDEX_CLEANUP: + errcontext(_("while cleaning up index \"%s.%s\" of relation \"%s\""), + cbarg->relnamespace, cbarg->indname, cbarg->relname); + break; + } +} -- 2.7.4 --byLs0wutDcxFdwtm Content-Type: text/x-diff; charset=us-ascii Content-Disposition: attachment; filename="v17-0002-Functions-to-initialize-errcontext.patch" ^ permalink raw reply [nested|flat] 10+ messages in thread
* [PATCH v22 1/3] vacuum errcontext to show block being processed @ 2019-12-13 02:54 Justin Pryzby <pryzbyj@telsasoft.com> 0 siblings, 0 replies; 10+ messages in thread From: Justin Pryzby @ 2019-12-13 02:54 UTC (permalink / raw) Discussion: https://www.postgresql.org/message-id/20191120210600.GC30362@telsasoft.com --- src/backend/access/heap/vacuumlazy.c | 162 +++++++++++++++++++++++++++++++---- 1 file changed, 144 insertions(+), 18 deletions(-) diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c index 03c43ef..d34d0c5 100644 --- a/src/backend/access/heap/vacuumlazy.c +++ b/src/backend/access/heap/vacuumlazy.c @@ -270,6 +270,8 @@ typedef struct LVParallelState typedef struct LVRelStats { + char *relnamespace; + char *relname; /* useindex = true means two-pass strategy; false means one-pass */ bool useindex; /* Overall statistics about rel */ @@ -290,8 +292,12 @@ typedef struct LVRelStats int num_index_scans; TransactionId latestRemovedXid; bool lock_waiter_detected; -} LVRelStats; + /* Used for error callback: */ + char *indname; + BlockNumber blkno; /* used only for heap operations */ + int phase; /* Reusing same constants as for progress reporting */ +} LVRelStats; /* A few variables that don't seem worth passing around as parameters */ static int elevel = -1; @@ -314,10 +320,10 @@ static void lazy_vacuum_all_indexes(Relation onerel, Relation *Irel, LVRelStats *vacrelstats, LVParallelState *lps, int nindexes); static void lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, - LVDeadTuples *dead_tuples, double reltuples); + LVDeadTuples *dead_tuples, double reltuples, LVRelStats *vacrelstats); static void lazy_cleanup_index(Relation indrel, IndexBulkDeleteResult **stats, - double reltuples, bool estimated_count); + double reltuples, bool estimated_count, LVRelStats *vacrelstats); static int lazy_vacuum_page(Relation onerel, BlockNumber blkno, Buffer buffer, int tupindex, LVRelStats *vacrelstats, Buffer *vmbuffer); static bool should_attempt_truncation(VacuumParams *params, @@ -361,6 +367,9 @@ static void end_parallel_vacuum(Relation *Irel, IndexBulkDeleteResult **stats, LVParallelState *lps, int nindexes); static LVSharedIndStats *get_indstats(LVShared *lvshared, int n); static bool skip_parallel_vacuum_index(Relation indrel, LVShared *lvshared); +static void vacuum_error_callback(void *arg); +static void update_vacuum_error_cbarg(LVRelStats *errcbarg, int phase, + BlockNumber blkno, Relation rel); /* @@ -460,6 +469,9 @@ heap_vacuum_rel(Relation onerel, VacuumParams *params, vacrelstats = (LVRelStats *) palloc0(sizeof(LVRelStats)); + vacrelstats->relnamespace = get_namespace_name(RelationGetNamespace(onerel)); + vacrelstats->relname = pstrdup(RelationGetRelationName(onerel)); + vacrelstats->indname = NULL; vacrelstats->old_rel_pages = onerel->rd_rel->relpages; vacrelstats->old_live_tuples = onerel->rd_rel->reltuples; vacrelstats->num_index_scans = 0; @@ -699,7 +711,6 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, BlockNumber nblocks, blkno; HeapTupleData tuple; - char *relname; TransactionId relfrozenxid = onerel->rd_rel->relfrozenxid; TransactionId relminmxid = onerel->rd_rel->relminmxid; BlockNumber empty_pages, @@ -724,20 +735,20 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, PROGRESS_VACUUM_MAX_DEAD_TUPLES }; int64 initprog_val[3]; + ErrorContextCallback errcallback; pg_rusage_init(&ru0); - relname = RelationGetRelationName(onerel); if (aggressive) ereport(elevel, (errmsg("aggressively vacuuming \"%s.%s\"", - get_namespace_name(RelationGetNamespace(onerel)), - relname))); + vacrelstats->relnamespace, + vacrelstats->relname))); else ereport(elevel, (errmsg("vacuuming \"%s.%s\"", - get_namespace_name(RelationGetNamespace(onerel)), - relname))); + vacrelstats->relnamespace, + vacrelstats->relname))); empty_pages = vacuumed_pages = 0; next_fsm_block_to_vacuum = (BlockNumber) 0; @@ -870,6 +881,14 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, else skipping_blocks = false; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, PROGRESS_VACUUM_PHASE_SCAN_HEAP, + InvalidBlockNumber, NULL); + errcallback.callback = vacuum_error_callback; + errcallback.arg = vacrelstats; + errcallback.previous = error_context_stack; + error_context_stack = &errcallback; + for (blkno = 0; blkno < nblocks; blkno++) { Buffer buf; @@ -891,6 +910,8 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, #define FORCE_CHECK_PAGE() \ (blkno == nblocks - 1 && should_attempt_truncation(params, vacrelstats)) + vacrelstats->blkno = blkno; + pgstat_progress_update_param(PROGRESS_VACUUM_HEAP_BLKS_SCANNED, blkno); if (blkno == next_unskippable_block) @@ -1011,6 +1032,10 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, /* Report that we are once again scanning the heap */ pgstat_progress_update_param(PROGRESS_VACUUM_PHASE, PROGRESS_VACUUM_PHASE_SCAN_HEAP); + + /* Set the error context while continuing heap scan */ + update_vacuum_error_cbarg(vacrelstats, + PROGRESS_VACUUM_PHASE_SCAN_HEAP, blkno, NULL); } /* @@ -1534,7 +1559,7 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, && VM_ALL_VISIBLE(onerel, blkno, &vmbuffer)) { elog(WARNING, "page is not marked all-visible but visibility map bit is set in relation \"%s\" page %u", - relname, blkno); + vacrelstats->relname, blkno); visibilitymap_clear(onerel, blkno, vmbuffer, VISIBILITYMAP_VALID_BITS); } @@ -1555,7 +1580,7 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, else if (PageIsAllVisible(page) && has_dead_tuples) { elog(WARNING, "page containing dead tuples is marked as all-visible in relation \"%s\" page %u", - relname, blkno); + vacrelstats->relname, blkno); PageClearAllVisible(page); MarkBufferDirty(buf); visibilitymap_clear(onerel, blkno, vmbuffer, @@ -1651,6 +1676,9 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, if (vacrelstats->useindex) lazy_cleanup_all_indexes(Irel, indstats, vacrelstats, lps, nindexes); + /* Pop the error context stack */ + error_context_stack = errcallback.previous; + /* * End parallel mode before updating index statistics as we cannot write * during parallel mode. @@ -1744,7 +1772,7 @@ lazy_vacuum_all_indexes(Relation onerel, Relation *Irel, for (idx = 0; idx < nindexes; idx++) lazy_vacuum_index(Irel[idx], &stats[idx], vacrelstats->dead_tuples, - vacrelstats->old_live_tuples); + vacrelstats->old_live_tuples, vacrelstats); } /* Increase and report the number of index scans */ @@ -1778,8 +1806,13 @@ lazy_vacuum_heap(Relation onerel, LVRelStats *vacrelstats) PROGRESS_VACUUM_PHASE_VACUUM_HEAP); pg_rusage_init(&ru0); - npages = 0; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, + PROGRESS_VACUUM_PHASE_VACUUM_HEAP, InvalidBlockNumber, + NULL); + + npages = 0; tupindex = 0; while (tupindex < vacrelstats->dead_tuples->num_tuples) { @@ -1791,6 +1824,7 @@ lazy_vacuum_heap(Relation onerel, LVRelStats *vacrelstats) vacuum_delay_point(); tblk = ItemPointerGetBlockNumber(&vacrelstats->dead_tuples->itemptrs[tupindex]); + vacrelstats->blkno = tblk; buf = ReadBufferExtended(onerel, MAIN_FORKNUM, tblk, RBM_NORMAL, vac_strategy); if (!ConditionalLockBufferForCleanup(buf)) @@ -2203,6 +2237,26 @@ vacuum_one_index(Relation indrel, IndexBulkDeleteResult **stats, LVDeadTuples *dead_tuples) { IndexBulkDeleteResult *bulkdelete_res = NULL; + /* Init vacrelstats for error callback by parallel workers: */ + LVRelStats vacrelstats = + { + .relnamespace = get_namespace_name(RelationGetNamespace(indrel)), + .relname = pstrdup(get_rel_name(indrel->rd_index->indrelid)), + .indname = pstrdup(RelationGetRelationName(indrel)), + }; + + /* Setup error traceback support for ereport() */ + ErrorContextCallback errcallback; + update_vacuum_error_cbarg(&vacrelstats, + lvshared->for_cleanup ? + PROGRESS_VACUUM_PHASE_INDEX_CLEANUP : + PROGRESS_VACUUM_PHASE_VACUUM_INDEX, + InvalidBlockNumber, NULL); + + errcallback.callback = vacuum_error_callback; + errcallback.arg = &vacrelstats; + errcallback.previous = error_context_stack; + error_context_stack = &errcallback; if (shared_indstats) { @@ -2220,10 +2274,13 @@ vacuum_one_index(Relation indrel, IndexBulkDeleteResult **stats, /* Do vacuum or cleanup of the index */ if (lvshared->for_cleanup) lazy_cleanup_index(indrel, stats, lvshared->reltuples, - lvshared->estimated_count); + lvshared->estimated_count, &vacrelstats); else lazy_vacuum_index(indrel, stats, dead_tuples, - lvshared->reltuples); + lvshared->reltuples, &vacrelstats); + + /* Pop the error context stack */ + error_context_stack = errcallback.previous; /* * Copy the index bulk-deletion result returned from ambulkdelete and @@ -2298,7 +2355,8 @@ lazy_cleanup_all_indexes(Relation *Irel, IndexBulkDeleteResult **stats, for (idx = 0; idx < nindexes; idx++) lazy_cleanup_index(Irel[idx], &stats[idx], vacrelstats->new_rel_tuples, - vacrelstats->tupcount_pages < vacrelstats->rel_pages); + vacrelstats->tupcount_pages < vacrelstats->rel_pages, + vacrelstats); } } @@ -2313,7 +2371,7 @@ lazy_cleanup_all_indexes(Relation *Irel, IndexBulkDeleteResult **stats, */ static void lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, - LVDeadTuples *dead_tuples, double reltuples) + LVDeadTuples *dead_tuples, double reltuples, LVRelStats *vacrelstats) { IndexVacuumInfo ivinfo; const char *msg; @@ -2329,6 +2387,11 @@ lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, ivinfo.num_heap_tuples = reltuples; ivinfo.strategy = vac_strategy; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, + PROGRESS_VACUUM_PHASE_VACUUM_INDEX, InvalidBlockNumber, + indrel); + /* Do bulk deletion */ *stats = index_bulk_delete(&ivinfo, *stats, lazy_tid_reaped, (void *) dead_tuples); @@ -2354,7 +2417,7 @@ lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, static void lazy_cleanup_index(Relation indrel, IndexBulkDeleteResult **stats, - double reltuples, bool estimated_count) + double reltuples, bool estimated_count, LVRelStats *vacrelstats) { IndexVacuumInfo ivinfo; const char *msg; @@ -2371,6 +2434,10 @@ lazy_cleanup_index(Relation indrel, ivinfo.num_heap_tuples = reltuples; ivinfo.strategy = vac_strategy; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, + PROGRESS_VACUUM_PHASE_INDEX_CLEANUP, InvalidBlockNumber, indrel); + *stats = index_vacuum_cleanup(&ivinfo, *stats); if (!(*stats)) @@ -3376,3 +3443,62 @@ parallel_vacuum_main(dsm_segment *seg, shm_toc *toc) table_close(onerel, ShareUpdateExclusiveLock); pfree(stats); } + +/* + * Error context callback for errors occurring during vacuum. + */ +static void +vacuum_error_callback(void *arg) +{ + LVRelStats *cbarg = arg; + + switch (cbarg->phase) { + case PROGRESS_VACUUM_PHASE_SCAN_HEAP: + if (BlockNumberIsValid(cbarg->blkno)) + errcontext("while scanning block %u of relation \"%s.%s\"", + cbarg->blkno, cbarg->relnamespace, cbarg->relname); + break; + + case PROGRESS_VACUUM_PHASE_VACUUM_HEAP: + if (BlockNumberIsValid(cbarg->blkno)) + errcontext("while vacuuming block %u of relation \"%s.%s\"", + cbarg->blkno, cbarg->relnamespace, cbarg->relname); + break; + + case PROGRESS_VACUUM_PHASE_VACUUM_INDEX: + errcontext("while vacuuming index \"%s\" of relation \"%s.%s\"", + cbarg->indname, cbarg->relnamespace, cbarg->relname); + break; + + case PROGRESS_VACUUM_PHASE_INDEX_CLEANUP: + errcontext("while cleaning up index \"%s\" of relation \"%s.%s\"", + cbarg->indname, cbarg->relnamespace, cbarg->relname); + break; + + case PROGRESS_VACUUM_PHASE_TRUNCATE: + case PROGRESS_VACUUM_PHASE_FINAL_CLEANUP: + default: + return; /* Shouldn't happen: do nothing */ + } +} + +/* Update vacuum error callback for current phase, block and index */ +static void +update_vacuum_error_cbarg(LVRelStats *errcbarg, int phase, BlockNumber blkno, + Relation indrel) +{ + errcbarg->blkno = blkno; + errcbarg->phase = phase; + + /* Free index name from any previous phase */ + if (errcbarg->indname) { + pfree(errcbarg->indname); + errcbarg->indname = NULL; + } + + /* For index phases, save the name of the current index for the callback */ + if (indrel) { + Assert(indrel->rd_rel->relkind == RELKIND_INDEX); + errcbarg->indname = pstrdup(RelationGetRelationName(indrel)); + } +} -- 2.7.4 --Q68bSM7Ycu6FN28Q Content-Type: text/x-diff; charset=us-ascii Content-Disposition: attachment; filename="v22-0002-add-callback-for-truncation.patch" ^ permalink raw reply [nested|flat] 10+ messages in thread
* [PATCH v9 2/3] vacuum errcontext to show block being processed @ 2019-12-13 02:54 Justin Pryzby <pryzbyj@telsasoft.com> 0 siblings, 0 replies; 10+ messages in thread From: Justin Pryzby @ 2019-12-13 02:54 UTC (permalink / raw) As requested here. https://www.postgresql.org/message-id/20190807235154.erbmr4o4bo6vgnjv%40alap3.anarazel.de --- src/backend/access/heap/vacuumlazy.c | 37 ++++++++++++++++++++++++++++++++++++ 1 file changed, 37 insertions(+) diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c index 9849685..c96abdf 100644 --- a/src/backend/access/heap/vacuumlazy.c +++ b/src/backend/access/heap/vacuumlazy.c @@ -289,6 +289,12 @@ typedef struct LVRelStats bool lock_waiter_detected; } LVRelStats; +typedef struct +{ + char *relname; + char *relnamespace; + BlockNumber blkno; +} vacuum_error_callback_arg; /* A few variables that don't seem worth passing around as parameters */ static int elevel = -1; @@ -358,6 +364,7 @@ static void end_parallel_vacuum(Relation *Irel, IndexBulkDeleteResult **stats, LVParallelState *lps, int nindexes); static LVSharedIndStats *get_indstats(LVShared *lvshared, int n); static bool skip_parallel_vacuum_index(Relation indrel, LVShared *lvshared); +static void vacuum_error_callback(void *arg); /* @@ -803,6 +810,8 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, PROGRESS_VACUUM_MAX_DEAD_TUPLES }; int64 initprog_val[3]; + ErrorContextCallback errcallback; + vacuum_error_callback_arg errcbarg; pg_rusage_init(&ru0); @@ -879,6 +888,15 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, next_unskippable_block = 0; skipping_blocks = skip_blocks(onerel, params, &next_unskippable_block, nblocks, &vmbuffer, aggressive); + /* Setup error traceback support for ereport() */ + errcbarg.relnamespace = get_namespace_name(RelationGetNamespace(onerel)); + errcbarg.relname = relname; + errcbarg.blkno = InvalidBlockNumber; /* Not known yet */ + errcallback.callback = vacuum_error_callback; + errcallback.arg = (void *) &errcbarg; + errcallback.previous = error_context_stack; + error_context_stack = &errcallback; + for (blkno = 0; blkno < nblocks; blkno++) { Buffer buf; @@ -900,6 +918,8 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, #define FORCE_CHECK_PAGE() \ (blkno == nblocks - 1 && should_attempt_truncation(params, vacrelstats)) + errcbarg.blkno = blkno; + pgstat_progress_update_param(PROGRESS_VACUUM_HEAP_BLKS_SCANNED, blkno); if (blkno == next_unskippable_block) @@ -966,8 +986,11 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, } /* Work on all the indexes, then the heap */ + /* Don't use the errcontext handler outside this function */ + error_context_stack = errcallback.previous; lazy_vacuum_all_indexes(onerel, Irel, indstats, vacrelstats, lps, nindexes); + error_context_stack = &errcallback; /* Remove tuples from heap */ lazy_vacuum_heap(onerel, vacrelstats); @@ -1575,6 +1598,9 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, RecordPageWithFreeSpace(onerel, blkno, freespace); } + /* Pop the error context stack */ + error_context_stack = errcallback.previous; + /* report that everything is scanned and vacuumed */ pgstat_progress_update_param(PROGRESS_VACUUM_HEAP_BLKS_SCANNED, blkno); @@ -3353,3 +3379,14 @@ parallel_vacuum_main(dsm_segment *seg, shm_toc *toc) table_close(onerel, ShareUpdateExclusiveLock); pfree(stats); } + +/* + * Error context callback for errors occurring during vacuum. + */ +static void +vacuum_error_callback(void *arg) +{ + vacuum_error_callback_arg *cbarg = arg; + errcontext("while scanning block %u of relation \"%s.%s\"", + cbarg->blkno, cbarg->relnamespace, cbarg->relname); +} -- 2.7.4 --LwgCY5hL6Igino+O Content-Type: text/x-diff; charset=us-ascii Content-Disposition: attachment; filename="v9-0003-add-errcontext-callback-in-lazy_vacuum_heap-too.patch" ^ permalink raw reply [nested|flat] 10+ messages in thread
* [PATCH v27 1/5] vacuum errcontext to show block being processed @ 2019-12-13 02:54 Justin Pryzby <pryzbyj@telsasoft.com> 0 siblings, 0 replies; 10+ messages in thread From: Justin Pryzby @ 2019-12-13 02:54 UTC (permalink / raw) Discussion: https://www.postgresql.org/message-id/20191120210600.GC30362@telsasoft.com --- src/backend/access/heap/vacuumlazy.c | 215 +++++++++++++++++++++++---- src/tools/pgindent/typedefs.list | 1 + 2 files changed, 191 insertions(+), 25 deletions(-) diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c index 03c43efc32..92bac9a24d 100644 --- a/src/backend/access/heap/vacuumlazy.c +++ b/src/backend/access/heap/vacuumlazy.c @@ -268,8 +268,19 @@ typedef struct LVParallelState int nindexes_parallel_condcleanup; } LVParallelState; +typedef enum +{ + VACUUM_ERRCB_PHASE_UNKNOWN, + VACUUM_ERRCB_PHASE_SCAN_HEAP, + VACUUM_ERRCB_PHASE_VACUUM_INDEX, + VACUUM_ERRCB_PHASE_VACUUM_HEAP, + VACUUM_ERRCB_PHASE_INDEX_CLEANUP, +} errcb_phase; + typedef struct LVRelStats { + char *relnamespace; + char *relname; /* useindex = true means two-pass strategy; false means one-pass */ bool useindex; /* Overall statistics about rel */ @@ -290,8 +301,12 @@ typedef struct LVRelStats int num_index_scans; TransactionId latestRemovedXid; bool lock_waiter_detected; -} LVRelStats; + /* Used for error callback: */ + char *indname; + BlockNumber blkno; /* used only for heap operations */ + errcb_phase phase; +} LVRelStats; /* A few variables that don't seem worth passing around as parameters */ static int elevel = -1; @@ -314,10 +329,10 @@ static void lazy_vacuum_all_indexes(Relation onerel, Relation *Irel, LVRelStats *vacrelstats, LVParallelState *lps, int nindexes); static void lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, - LVDeadTuples *dead_tuples, double reltuples); + LVDeadTuples *dead_tuples, double reltuples, LVRelStats *vacrelstats); static void lazy_cleanup_index(Relation indrel, IndexBulkDeleteResult **stats, - double reltuples, bool estimated_count); + double reltuples, bool estimated_count, LVRelStats *vacrelstats); static int lazy_vacuum_page(Relation onerel, BlockNumber blkno, Buffer buffer, int tupindex, LVRelStats *vacrelstats, Buffer *vmbuffer); static bool should_attempt_truncation(VacuumParams *params, @@ -337,13 +352,13 @@ static void lazy_parallel_vacuum_indexes(Relation *Irel, IndexBulkDeleteResult * int nindexes); static void parallel_vacuum_index(Relation *Irel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVDeadTuples *dead_tuples, - int nindexes); + int nindexes, LVRelStats *vacrelstats); static void vacuum_indexes_leader(Relation *Irel, IndexBulkDeleteResult **stats, LVRelStats *vacrelstats, LVParallelState *lps, int nindexes); static void vacuum_one_index(Relation indrel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVSharedIndStats *shared_indstats, - LVDeadTuples *dead_tuples); + LVDeadTuples *dead_tuples, LVRelStats *vacrelstats); static void lazy_cleanup_all_indexes(Relation *Irel, IndexBulkDeleteResult **stats, LVRelStats *vacrelstats, LVParallelState *lps, int nindexes); @@ -361,6 +376,9 @@ static void end_parallel_vacuum(Relation *Irel, IndexBulkDeleteResult **stats, LVParallelState *lps, int nindexes); static LVSharedIndStats *get_indstats(LVShared *lvshared, int n); static bool skip_parallel_vacuum_index(Relation indrel, LVShared *lvshared); +static void vacuum_error_callback(void *arg); +static void update_vacuum_error_cbarg(LVRelStats *errcbarg, int phase, + BlockNumber blkno, Relation rel); /* @@ -460,6 +478,9 @@ heap_vacuum_rel(Relation onerel, VacuumParams *params, vacrelstats = (LVRelStats *) palloc0(sizeof(LVRelStats)); + vacrelstats->relnamespace = get_namespace_name(RelationGetNamespace(onerel)); + vacrelstats->relname = pstrdup(RelationGetRelationName(onerel)); + vacrelstats->indname = NULL; vacrelstats->old_rel_pages = onerel->rd_rel->relpages; vacrelstats->old_live_tuples = onerel->rd_rel->reltuples; vacrelstats->num_index_scans = 0; @@ -699,7 +720,6 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, BlockNumber nblocks, blkno; HeapTupleData tuple; - char *relname; TransactionId relfrozenxid = onerel->rd_rel->relfrozenxid; TransactionId relminmxid = onerel->rd_rel->relminmxid; BlockNumber empty_pages, @@ -724,20 +744,20 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, PROGRESS_VACUUM_MAX_DEAD_TUPLES }; int64 initprog_val[3]; + ErrorContextCallback errcallback; pg_rusage_init(&ru0); - relname = RelationGetRelationName(onerel); if (aggressive) ereport(elevel, (errmsg("aggressively vacuuming \"%s.%s\"", - get_namespace_name(RelationGetNamespace(onerel)), - relname))); + vacrelstats->relnamespace, + vacrelstats->relname))); else ereport(elevel, (errmsg("vacuuming \"%s.%s\"", - get_namespace_name(RelationGetNamespace(onerel)), - relname))); + vacrelstats->relnamespace, + vacrelstats->relname))); empty_pages = vacuumed_pages = 0; next_fsm_block_to_vacuum = (BlockNumber) 0; @@ -870,6 +890,14 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, else skipping_blocks = false; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, VACUUM_ERRCB_PHASE_SCAN_HEAP, + InvalidBlockNumber, NULL); + errcallback.callback = vacuum_error_callback; + errcallback.arg = vacrelstats; + errcallback.previous = error_context_stack; + error_context_stack = &errcallback; + for (blkno = 0; blkno < nblocks; blkno++) { Buffer buf; @@ -891,8 +919,13 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, #define FORCE_CHECK_PAGE() \ (blkno == nblocks - 1 && should_attempt_truncation(params, vacrelstats)) + vacrelstats->blkno = blkno; + pgstat_progress_update_param(PROGRESS_VACUUM_HEAP_BLKS_SCANNED, blkno); + update_vacuum_error_cbarg(vacrelstats, VACUUM_ERRCB_PHASE_SCAN_HEAP, + blkno, NULL); + if (blkno == next_unskippable_block) { /* Time to advance next_unskippable_block */ @@ -1011,6 +1044,10 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, /* Report that we are once again scanning the heap */ pgstat_progress_update_param(PROGRESS_VACUUM_PHASE, PROGRESS_VACUUM_PHASE_SCAN_HEAP); + + /* Set the error context while continuing heap scan */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_SCAN_HEAP, blkno, NULL); } /* @@ -1492,6 +1529,11 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, blkno); next_fsm_block_to_vacuum = blkno; } + + /* Set the error context while continuing heap scan */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_SCAN_HEAP, blkno, NULL); + } freespace = PageGetHeapFreeSpace(page); @@ -1534,7 +1576,7 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, && VM_ALL_VISIBLE(onerel, blkno, &vmbuffer)) { elog(WARNING, "page is not marked all-visible but visibility map bit is set in relation \"%s\" page %u", - relname, blkno); + vacrelstats->relname, blkno); visibilitymap_clear(onerel, blkno, vmbuffer, VISIBILITYMAP_VALID_BITS); } @@ -1555,7 +1597,7 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, else if (PageIsAllVisible(page) && has_dead_tuples) { elog(WARNING, "page containing dead tuples is marked as all-visible in relation \"%s\" page %u", - relname, blkno); + vacrelstats->relname, blkno); PageClearAllVisible(page); MarkBufferDirty(buf); visibilitymap_clear(onerel, blkno, vmbuffer, @@ -1651,6 +1693,9 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, if (vacrelstats->useindex) lazy_cleanup_all_indexes(Irel, indstats, vacrelstats, lps, nindexes); + /* Pop the error context stack */ + error_context_stack = errcallback.previous; + /* * End parallel mode before updating index statistics as we cannot write * during parallel mode. @@ -1744,7 +1789,7 @@ lazy_vacuum_all_indexes(Relation onerel, Relation *Irel, for (idx = 0; idx < nindexes; idx++) lazy_vacuum_index(Irel[idx], &stats[idx], vacrelstats->dead_tuples, - vacrelstats->old_live_tuples); + vacrelstats->old_live_tuples, vacrelstats); } /* Increase and report the number of index scans */ @@ -1777,6 +1822,10 @@ lazy_vacuum_heap(Relation onerel, LVRelStats *vacrelstats) pgstat_progress_update_param(PROGRESS_VACUUM_PHASE, PROGRESS_VACUUM_PHASE_VACUUM_HEAP); + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, VACUUM_ERRCB_PHASE_VACUUM_HEAP, + InvalidBlockNumber, NULL); + pg_rusage_init(&ru0); npages = 0; @@ -1791,6 +1840,7 @@ lazy_vacuum_heap(Relation onerel, LVRelStats *vacrelstats) vacuum_delay_point(); tblk = ItemPointerGetBlockNumber(&vacrelstats->dead_tuples->itemptrs[tupindex]); + vacrelstats->blkno = tblk; buf = ReadBufferExtended(onerel, MAIN_FORKNUM, tblk, RBM_NORMAL, vac_strategy); if (!ConditionalLockBufferForCleanup(buf)) @@ -1822,6 +1872,11 @@ lazy_vacuum_heap(Relation onerel, LVRelStats *vacrelstats) RelationGetRelationName(onerel), tupindex, npages), errdetail_internal("%s", pg_rusage_show(&ru0)))); + + /* Clear the error traceback phase */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_UNKNOWN, InvalidBlockNumber, + NULL); } /* @@ -1847,6 +1902,10 @@ lazy_vacuum_page(Relation onerel, BlockNumber blkno, Buffer buffer, pgstat_progress_update_param(PROGRESS_VACUUM_HEAP_BLKS_VACUUMED, blkno); + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, VACUUM_ERRCB_PHASE_VACUUM_HEAP, + blkno, NULL); + START_CRIT_SECTION(); for (; tupindex < dead_tuples->num_tuples; tupindex++) @@ -1923,6 +1982,11 @@ lazy_vacuum_page(Relation onerel, BlockNumber blkno, Buffer buffer, *vmbuffer, visibility_cutoff_xid, flags); } + /* Clear the error traceback phase */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_UNKNOWN, InvalidBlockNumber, + NULL); + return tupindex; } @@ -2083,7 +2147,7 @@ lazy_parallel_vacuum_indexes(Relation *Irel, IndexBulkDeleteResult **stats, * indexes in the case where no workers are launched. */ parallel_vacuum_index(Irel, stats, lps->lvshared, - vacrelstats->dead_tuples, nindexes); + vacrelstats->dead_tuples, nindexes, vacrelstats); /* Wait for all vacuum workers to finish */ WaitForParallelWorkersToFinish(lps->pcxt); @@ -2106,7 +2170,7 @@ lazy_parallel_vacuum_indexes(Relation *Irel, IndexBulkDeleteResult **stats, static void parallel_vacuum_index(Relation *Irel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVDeadTuples *dead_tuples, - int nindexes) + int nindexes, LVRelStats *vacrelstats) { /* * Increment the active worker count if we are able to launch any worker. @@ -2140,7 +2204,7 @@ parallel_vacuum_index(Relation *Irel, IndexBulkDeleteResult **stats, /* Do vacuum or cleanup of the index */ vacuum_one_index(Irel[idx], &(stats[idx]), lvshared, shared_indstats, - dead_tuples); + dead_tuples, vacrelstats); } /* @@ -2180,7 +2244,8 @@ vacuum_indexes_leader(Relation *Irel, IndexBulkDeleteResult **stats, if (shared_indstats == NULL || skip_parallel_vacuum_index(Irel[i], lps->lvshared)) vacuum_one_index(Irel[i], &(stats[i]), lps->lvshared, - shared_indstats, vacrelstats->dead_tuples); + shared_indstats, vacrelstats->dead_tuples, + vacrelstats); } /* @@ -2200,7 +2265,7 @@ vacuum_indexes_leader(Relation *Irel, IndexBulkDeleteResult **stats, static void vacuum_one_index(Relation indrel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVSharedIndStats *shared_indstats, - LVDeadTuples *dead_tuples) + LVDeadTuples *dead_tuples, LVRelStats *vacrelstats) { IndexBulkDeleteResult *bulkdelete_res = NULL; @@ -2220,10 +2285,10 @@ vacuum_one_index(Relation indrel, IndexBulkDeleteResult **stats, /* Do vacuum or cleanup of the index */ if (lvshared->for_cleanup) lazy_cleanup_index(indrel, stats, lvshared->reltuples, - lvshared->estimated_count); + lvshared->estimated_count, vacrelstats); else lazy_vacuum_index(indrel, stats, dead_tuples, - lvshared->reltuples); + lvshared->reltuples, vacrelstats); /* * Copy the index bulk-deletion result returned from ambulkdelete and @@ -2298,7 +2363,8 @@ lazy_cleanup_all_indexes(Relation *Irel, IndexBulkDeleteResult **stats, for (idx = 0; idx < nindexes; idx++) lazy_cleanup_index(Irel[idx], &stats[idx], vacrelstats->new_rel_tuples, - vacrelstats->tupcount_pages < vacrelstats->rel_pages); + vacrelstats->tupcount_pages < vacrelstats->rel_pages, + vacrelstats); } } @@ -2313,7 +2379,7 @@ lazy_cleanup_all_indexes(Relation *Irel, IndexBulkDeleteResult **stats, */ static void lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, - LVDeadTuples *dead_tuples, double reltuples) + LVDeadTuples *dead_tuples, double reltuples, LVRelStats *vacrelstats) { IndexVacuumInfo ivinfo; const char *msg; @@ -2329,6 +2395,11 @@ lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, ivinfo.num_heap_tuples = reltuples; ivinfo.strategy = vac_strategy; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_VACUUM_INDEX, InvalidBlockNumber, + indrel); + /* Do bulk deletion */ *stats = index_bulk_delete(&ivinfo, *stats, lazy_tid_reaped, (void *) dead_tuples); @@ -2343,6 +2414,11 @@ lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, RelationGetRelationName(indrel), dead_tuples->num_tuples), errdetail_internal("%s", pg_rusage_show(&ru0)))); + + /* Clear the error traceback phase */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_UNKNOWN, InvalidBlockNumber, + NULL); } /* @@ -2354,7 +2430,7 @@ lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, static void lazy_cleanup_index(Relation indrel, IndexBulkDeleteResult **stats, - double reltuples, bool estimated_count) + double reltuples, bool estimated_count, LVRelStats *vacrelstats) { IndexVacuumInfo ivinfo; const char *msg; @@ -2371,6 +2447,10 @@ lazy_cleanup_index(Relation indrel, ivinfo.num_heap_tuples = reltuples; ivinfo.strategy = vac_strategy; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_INDEX_CLEANUP, InvalidBlockNumber, indrel); + *stats = index_vacuum_cleanup(&ivinfo, *stats); if (!(*stats)) @@ -2392,6 +2472,11 @@ lazy_cleanup_index(Relation indrel, (*stats)->tuples_removed, (*stats)->pages_deleted, (*stats)->pages_free, pg_rusage_show(&ru0)))); + + /* Clear the error traceback phase */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_UNKNOWN, InvalidBlockNumber, + NULL); } /* @@ -3320,6 +3405,8 @@ parallel_vacuum_main(dsm_segment *seg, shm_toc *toc) int nindexes; char *sharedquery; IndexBulkDeleteResult **stats; + LVRelStats vacrelstats; + ErrorContextCallback errcallback; lvshared = (LVShared *) shm_toc_lookup(toc, PARALLEL_VACUUM_KEY_SHARED, false); @@ -3369,10 +3456,88 @@ parallel_vacuum_main(dsm_segment *seg, shm_toc *toc) if (lvshared->maintenance_work_mem_worker > 0) maintenance_work_mem = lvshared->maintenance_work_mem_worker; + /* Init vacrelstats for use as error callback arg by parallel worker */ + vacrelstats.relnamespace = get_namespace_name(RelationGetNamespace(onerel)); + vacrelstats.relname = pstrdup(RelationGetRelationName(onerel)); + vacrelstats.indname = NULL; + vacrelstats.phase = VACUUM_ERRCB_PHASE_UNKNOWN; /* Not yet processing */ + + /* Setup error traceback support for ereport() */ + errcallback.callback = vacuum_error_callback; + errcallback.arg = &vacrelstats; + errcallback.previous = error_context_stack; + error_context_stack = &errcallback; + /* Process indexes to perform vacuum/cleanup */ - parallel_vacuum_index(indrels, stats, lvshared, dead_tuples, nindexes); + parallel_vacuum_index(indrels, stats, lvshared, dead_tuples, nindexes, + &vacrelstats); + + /* Pop the error context stack */ + error_context_stack = errcallback.previous; vac_close_indexes(nindexes, indrels, RowExclusiveLock); table_close(onerel, ShareUpdateExclusiveLock); pfree(stats); } + +/* + * Error context callback for errors occurring during vacuum. + */ +static void +vacuum_error_callback(void *arg) +{ + LVRelStats *cbarg = arg; + + switch (cbarg->phase) + { + case VACUUM_ERRCB_PHASE_SCAN_HEAP: + if (BlockNumberIsValid(cbarg->blkno)) + errcontext("while scanning block %u of relation \"%s.%s\"", + cbarg->blkno, cbarg->relnamespace, cbarg->relname); + break; + + case VACUUM_ERRCB_PHASE_VACUUM_HEAP: + if (BlockNumberIsValid(cbarg->blkno)) + errcontext("while vacuuming block %u of relation \"%s.%s\"", + cbarg->blkno, cbarg->relnamespace, cbarg->relname); + break; + + case VACUUM_ERRCB_PHASE_VACUUM_INDEX: + errcontext("while vacuuming index \"%s\" of relation \"%s.%s\"", + cbarg->indname, cbarg->relnamespace, cbarg->relname); + break; + + case VACUUM_ERRCB_PHASE_INDEX_CLEANUP: + errcontext("while cleaning up index \"%s\" of relation \"%s.%s\"", + cbarg->indname, cbarg->relnamespace, cbarg->relname); + break; + + case VACUUM_ERRCB_PHASE_UNKNOWN: + default: + return; /* do nothing; the cbarg may not be + * initialized */ + } +} + +/* Update vacuum error callback for current phase, block and index */ +static void +update_vacuum_error_cbarg(LVRelStats *errcbarg, int phase, BlockNumber blkno, + Relation indrel) +{ + errcbarg->blkno = blkno; + errcbarg->phase = phase; + + /* Free index name from any previous phase */ + if (errcbarg->indname) + { + pfree(errcbarg->indname); + errcbarg->indname = NULL; + } + + /* For index phases, save the name of the current index for the callback */ + if (indrel) + { + Assert(indrel->rd_rel->relkind == RELKIND_INDEX); + errcbarg->indname = pstrdup(RelationGetRelationName(indrel)); + } +} diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index fcfcf56f4f..1dee4e1ff2 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -582,6 +582,7 @@ EphemeralNamedRelationMetadata EphemeralNamedRelationMetadataData EquivalenceClass EquivalenceMember +errcb_phase ErrorContextCallback ErrorData EstimateDSMForeignScan_function -- 2.17.0 --gPQW1Pk7T/0rhUBV Content-Type: text/x-diff; charset=us-ascii Content-Disposition: attachment; filename="v27-0002-Save-the-values-of-the-callback.patch" ^ permalink raw reply [nested|flat] 10+ messages in thread
* [PATCH v28 1/5] vacuum errcontext to show block being processed @ 2019-12-13 02:54 Justin Pryzby <pryzbyj@telsasoft.com> 0 siblings, 0 replies; 10+ messages in thread From: Justin Pryzby @ 2019-12-13 02:54 UTC (permalink / raw) Discussion: https://www.postgresql.org/message-id/20191120210600.GC30362@telsasoft.com --- src/backend/access/heap/vacuumlazy.c | 209 +++++++++++++++++++++++---- src/tools/pgindent/typedefs.list | 1 + 2 files changed, 185 insertions(+), 25 deletions(-) diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c index 03c43efc32..768a69120b 100644 --- a/src/backend/access/heap/vacuumlazy.c +++ b/src/backend/access/heap/vacuumlazy.c @@ -268,8 +268,19 @@ typedef struct LVParallelState int nindexes_parallel_condcleanup; } LVParallelState; +typedef enum +{ + VACUUM_ERRCB_PHASE_UNKNOWN, + VACUUM_ERRCB_PHASE_SCAN_HEAP, + VACUUM_ERRCB_PHASE_VACUUM_INDEX, + VACUUM_ERRCB_PHASE_VACUUM_HEAP, + VACUUM_ERRCB_PHASE_INDEX_CLEANUP, +} errcb_phase; + typedef struct LVRelStats { + char *relnamespace; + char *relname; /* useindex = true means two-pass strategy; false means one-pass */ bool useindex; /* Overall statistics about rel */ @@ -290,8 +301,12 @@ typedef struct LVRelStats int num_index_scans; TransactionId latestRemovedXid; bool lock_waiter_detected; -} LVRelStats; + /* Used for error callback: */ + char *indname; + BlockNumber blkno; /* used only for heap operations */ + errcb_phase phase; +} LVRelStats; /* A few variables that don't seem worth passing around as parameters */ static int elevel = -1; @@ -314,10 +329,10 @@ static void lazy_vacuum_all_indexes(Relation onerel, Relation *Irel, LVRelStats *vacrelstats, LVParallelState *lps, int nindexes); static void lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, - LVDeadTuples *dead_tuples, double reltuples); + LVDeadTuples *dead_tuples, double reltuples, LVRelStats *vacrelstats); static void lazy_cleanup_index(Relation indrel, IndexBulkDeleteResult **stats, - double reltuples, bool estimated_count); + double reltuples, bool estimated_count, LVRelStats *vacrelstats); static int lazy_vacuum_page(Relation onerel, BlockNumber blkno, Buffer buffer, int tupindex, LVRelStats *vacrelstats, Buffer *vmbuffer); static bool should_attempt_truncation(VacuumParams *params, @@ -337,13 +352,13 @@ static void lazy_parallel_vacuum_indexes(Relation *Irel, IndexBulkDeleteResult * int nindexes); static void parallel_vacuum_index(Relation *Irel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVDeadTuples *dead_tuples, - int nindexes); + int nindexes, LVRelStats *vacrelstats); static void vacuum_indexes_leader(Relation *Irel, IndexBulkDeleteResult **stats, LVRelStats *vacrelstats, LVParallelState *lps, int nindexes); static void vacuum_one_index(Relation indrel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVSharedIndStats *shared_indstats, - LVDeadTuples *dead_tuples); + LVDeadTuples *dead_tuples, LVRelStats *vacrelstats); static void lazy_cleanup_all_indexes(Relation *Irel, IndexBulkDeleteResult **stats, LVRelStats *vacrelstats, LVParallelState *lps, int nindexes); @@ -361,6 +376,9 @@ static void end_parallel_vacuum(Relation *Irel, IndexBulkDeleteResult **stats, LVParallelState *lps, int nindexes); static LVSharedIndStats *get_indstats(LVShared *lvshared, int n); static bool skip_parallel_vacuum_index(Relation indrel, LVShared *lvshared); +static void vacuum_error_callback(void *arg); +static void update_vacuum_error_cbarg(LVRelStats *errcbarg, int phase, + BlockNumber blkno, char *indname); /* @@ -460,6 +478,9 @@ heap_vacuum_rel(Relation onerel, VacuumParams *params, vacrelstats = (LVRelStats *) palloc0(sizeof(LVRelStats)); + vacrelstats->relnamespace = get_namespace_name(RelationGetNamespace(onerel)); + vacrelstats->relname = pstrdup(RelationGetRelationName(onerel)); + vacrelstats->indname = NULL; vacrelstats->old_rel_pages = onerel->rd_rel->relpages; vacrelstats->old_live_tuples = onerel->rd_rel->reltuples; vacrelstats->num_index_scans = 0; @@ -699,7 +720,6 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, BlockNumber nblocks, blkno; HeapTupleData tuple; - char *relname; TransactionId relfrozenxid = onerel->rd_rel->relfrozenxid; TransactionId relminmxid = onerel->rd_rel->relminmxid; BlockNumber empty_pages, @@ -724,20 +744,20 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, PROGRESS_VACUUM_MAX_DEAD_TUPLES }; int64 initprog_val[3]; + ErrorContextCallback errcallback; pg_rusage_init(&ru0); - relname = RelationGetRelationName(onerel); if (aggressive) ereport(elevel, (errmsg("aggressively vacuuming \"%s.%s\"", - get_namespace_name(RelationGetNamespace(onerel)), - relname))); + vacrelstats->relnamespace, + vacrelstats->relname))); else ereport(elevel, (errmsg("vacuuming \"%s.%s\"", - get_namespace_name(RelationGetNamespace(onerel)), - relname))); + vacrelstats->relnamespace, + vacrelstats->relname))); empty_pages = vacuumed_pages = 0; next_fsm_block_to_vacuum = (BlockNumber) 0; @@ -870,6 +890,14 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, else skipping_blocks = false; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, VACUUM_ERRCB_PHASE_SCAN_HEAP, + InvalidBlockNumber, NULL); + errcallback.callback = vacuum_error_callback; + errcallback.arg = vacrelstats; + errcallback.previous = error_context_stack; + error_context_stack = &errcallback; + for (blkno = 0; blkno < nblocks; blkno++) { Buffer buf; @@ -891,8 +919,13 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, #define FORCE_CHECK_PAGE() \ (blkno == nblocks - 1 && should_attempt_truncation(params, vacrelstats)) + vacrelstats->blkno = blkno; + pgstat_progress_update_param(PROGRESS_VACUUM_HEAP_BLKS_SCANNED, blkno); + update_vacuum_error_cbarg(vacrelstats, VACUUM_ERRCB_PHASE_SCAN_HEAP, + blkno, NULL); + if (blkno == next_unskippable_block) { /* Time to advance next_unskippable_block */ @@ -1011,6 +1044,10 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, /* Report that we are once again scanning the heap */ pgstat_progress_update_param(PROGRESS_VACUUM_PHASE, PROGRESS_VACUUM_PHASE_SCAN_HEAP); + + /* Set the error context while continuing heap scan */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_SCAN_HEAP, blkno, NULL); } /* @@ -1457,6 +1494,10 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, lazy_vacuum_page(onerel, blkno, buf, 0, vacrelstats, &vmbuffer); vacuumed_pages++; has_dead_tuples = false; + + /* Set the error context while continuing heap scan */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_SCAN_HEAP, blkno, NULL); } else { @@ -1534,7 +1575,7 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, && VM_ALL_VISIBLE(onerel, blkno, &vmbuffer)) { elog(WARNING, "page is not marked all-visible but visibility map bit is set in relation \"%s\" page %u", - relname, blkno); + vacrelstats->relname, blkno); visibilitymap_clear(onerel, blkno, vmbuffer, VISIBILITYMAP_VALID_BITS); } @@ -1555,7 +1596,7 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, else if (PageIsAllVisible(page) && has_dead_tuples) { elog(WARNING, "page containing dead tuples is marked as all-visible in relation \"%s\" page %u", - relname, blkno); + vacrelstats->relname, blkno); PageClearAllVisible(page); MarkBufferDirty(buf); visibilitymap_clear(onerel, blkno, vmbuffer, @@ -1651,6 +1692,9 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, if (vacrelstats->useindex) lazy_cleanup_all_indexes(Irel, indstats, vacrelstats, lps, nindexes); + /* Pop the error context stack */ + error_context_stack = errcallback.previous; + /* * End parallel mode before updating index statistics as we cannot write * during parallel mode. @@ -1744,7 +1788,7 @@ lazy_vacuum_all_indexes(Relation onerel, Relation *Irel, for (idx = 0; idx < nindexes; idx++) lazy_vacuum_index(Irel[idx], &stats[idx], vacrelstats->dead_tuples, - vacrelstats->old_live_tuples); + vacrelstats->old_live_tuples, vacrelstats); } /* Increase and report the number of index scans */ @@ -1777,6 +1821,10 @@ lazy_vacuum_heap(Relation onerel, LVRelStats *vacrelstats) pgstat_progress_update_param(PROGRESS_VACUUM_PHASE, PROGRESS_VACUUM_PHASE_VACUUM_HEAP); + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, VACUUM_ERRCB_PHASE_VACUUM_HEAP, + InvalidBlockNumber, NULL); + pg_rusage_init(&ru0); npages = 0; @@ -1791,6 +1839,7 @@ lazy_vacuum_heap(Relation onerel, LVRelStats *vacrelstats) vacuum_delay_point(); tblk = ItemPointerGetBlockNumber(&vacrelstats->dead_tuples->itemptrs[tupindex]); + vacrelstats->blkno = tblk; buf = ReadBufferExtended(onerel, MAIN_FORKNUM, tblk, RBM_NORMAL, vac_strategy); if (!ConditionalLockBufferForCleanup(buf)) @@ -1822,6 +1871,11 @@ lazy_vacuum_heap(Relation onerel, LVRelStats *vacrelstats) RelationGetRelationName(onerel), tupindex, npages), errdetail_internal("%s", pg_rusage_show(&ru0)))); + + /* Clear the error traceback phase */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_UNKNOWN, InvalidBlockNumber, + NULL); } /* @@ -1847,6 +1901,10 @@ lazy_vacuum_page(Relation onerel, BlockNumber blkno, Buffer buffer, pgstat_progress_update_param(PROGRESS_VACUUM_HEAP_BLKS_VACUUMED, blkno); + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, VACUUM_ERRCB_PHASE_VACUUM_HEAP, + blkno, NULL); + START_CRIT_SECTION(); for (; tupindex < dead_tuples->num_tuples; tupindex++) @@ -1923,6 +1981,11 @@ lazy_vacuum_page(Relation onerel, BlockNumber blkno, Buffer buffer, *vmbuffer, visibility_cutoff_xid, flags); } + /* Clear the error traceback phase */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_UNKNOWN, InvalidBlockNumber, + NULL); + return tupindex; } @@ -2083,7 +2146,7 @@ lazy_parallel_vacuum_indexes(Relation *Irel, IndexBulkDeleteResult **stats, * indexes in the case where no workers are launched. */ parallel_vacuum_index(Irel, stats, lps->lvshared, - vacrelstats->dead_tuples, nindexes); + vacrelstats->dead_tuples, nindexes, vacrelstats); /* Wait for all vacuum workers to finish */ WaitForParallelWorkersToFinish(lps->pcxt); @@ -2106,7 +2169,7 @@ lazy_parallel_vacuum_indexes(Relation *Irel, IndexBulkDeleteResult **stats, static void parallel_vacuum_index(Relation *Irel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVDeadTuples *dead_tuples, - int nindexes) + int nindexes, LVRelStats *vacrelstats) { /* * Increment the active worker count if we are able to launch any worker. @@ -2140,7 +2203,7 @@ parallel_vacuum_index(Relation *Irel, IndexBulkDeleteResult **stats, /* Do vacuum or cleanup of the index */ vacuum_one_index(Irel[idx], &(stats[idx]), lvshared, shared_indstats, - dead_tuples); + dead_tuples, vacrelstats); } /* @@ -2180,7 +2243,8 @@ vacuum_indexes_leader(Relation *Irel, IndexBulkDeleteResult **stats, if (shared_indstats == NULL || skip_parallel_vacuum_index(Irel[i], lps->lvshared)) vacuum_one_index(Irel[i], &(stats[i]), lps->lvshared, - shared_indstats, vacrelstats->dead_tuples); + shared_indstats, vacrelstats->dead_tuples, + vacrelstats); } /* @@ -2200,7 +2264,7 @@ vacuum_indexes_leader(Relation *Irel, IndexBulkDeleteResult **stats, static void vacuum_one_index(Relation indrel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVSharedIndStats *shared_indstats, - LVDeadTuples *dead_tuples) + LVDeadTuples *dead_tuples, LVRelStats *vacrelstats) { IndexBulkDeleteResult *bulkdelete_res = NULL; @@ -2220,10 +2284,10 @@ vacuum_one_index(Relation indrel, IndexBulkDeleteResult **stats, /* Do vacuum or cleanup of the index */ if (lvshared->for_cleanup) lazy_cleanup_index(indrel, stats, lvshared->reltuples, - lvshared->estimated_count); + lvshared->estimated_count, vacrelstats); else lazy_vacuum_index(indrel, stats, dead_tuples, - lvshared->reltuples); + lvshared->reltuples, vacrelstats); /* * Copy the index bulk-deletion result returned from ambulkdelete and @@ -2298,7 +2362,8 @@ lazy_cleanup_all_indexes(Relation *Irel, IndexBulkDeleteResult **stats, for (idx = 0; idx < nindexes; idx++) lazy_cleanup_index(Irel[idx], &stats[idx], vacrelstats->new_rel_tuples, - vacrelstats->tupcount_pages < vacrelstats->rel_pages); + vacrelstats->tupcount_pages < vacrelstats->rel_pages, + vacrelstats); } } @@ -2313,7 +2378,7 @@ lazy_cleanup_all_indexes(Relation *Irel, IndexBulkDeleteResult **stats, */ static void lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, - LVDeadTuples *dead_tuples, double reltuples) + LVDeadTuples *dead_tuples, double reltuples, LVRelStats *vacrelstats) { IndexVacuumInfo ivinfo; const char *msg; @@ -2329,6 +2394,11 @@ lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, ivinfo.num_heap_tuples = reltuples; ivinfo.strategy = vac_strategy; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_VACUUM_INDEX, InvalidBlockNumber, + RelationGetRelationName(indrel)); + /* Do bulk deletion */ *stats = index_bulk_delete(&ivinfo, *stats, lazy_tid_reaped, (void *) dead_tuples); @@ -2343,6 +2413,11 @@ lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, RelationGetRelationName(indrel), dead_tuples->num_tuples), errdetail_internal("%s", pg_rusage_show(&ru0)))); + + /* Clear the error traceback phase */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_UNKNOWN, InvalidBlockNumber, + NULL); } /* @@ -2354,7 +2429,7 @@ lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, static void lazy_cleanup_index(Relation indrel, IndexBulkDeleteResult **stats, - double reltuples, bool estimated_count) + double reltuples, bool estimated_count, LVRelStats *vacrelstats) { IndexVacuumInfo ivinfo; const char *msg; @@ -2371,6 +2446,12 @@ lazy_cleanup_index(Relation indrel, ivinfo.num_heap_tuples = reltuples; ivinfo.strategy = vac_strategy; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_INDEX_CLEANUP, + InvalidBlockNumber, + RelationGetRelationName(indrel)); + *stats = index_vacuum_cleanup(&ivinfo, *stats); if (!(*stats)) @@ -2392,6 +2473,11 @@ lazy_cleanup_index(Relation indrel, (*stats)->tuples_removed, (*stats)->pages_deleted, (*stats)->pages_free, pg_rusage_show(&ru0)))); + + /* Clear the error traceback phase */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_UNKNOWN, InvalidBlockNumber, + NULL); } /* @@ -3320,6 +3406,8 @@ parallel_vacuum_main(dsm_segment *seg, shm_toc *toc) int nindexes; char *sharedquery; IndexBulkDeleteResult **stats; + LVRelStats vacrelstats; + ErrorContextCallback errcallback; lvshared = (LVShared *) shm_toc_lookup(toc, PARALLEL_VACUUM_KEY_SHARED, false); @@ -3369,10 +3457,81 @@ parallel_vacuum_main(dsm_segment *seg, shm_toc *toc) if (lvshared->maintenance_work_mem_worker > 0) maintenance_work_mem = lvshared->maintenance_work_mem_worker; + /* Init vacrelstats for use as error callback arg by parallel worker */ + vacrelstats.relnamespace = get_namespace_name(RelationGetNamespace(onerel)); + vacrelstats.relname = pstrdup(RelationGetRelationName(onerel)); + vacrelstats.indname = NULL; + vacrelstats.phase = VACUUM_ERRCB_PHASE_UNKNOWN; /* Not yet processing */ + + /* Setup error traceback support for ereport() */ + errcallback.callback = vacuum_error_callback; + errcallback.arg = &vacrelstats; + errcallback.previous = error_context_stack; + error_context_stack = &errcallback; + /* Process indexes to perform vacuum/cleanup */ - parallel_vacuum_index(indrels, stats, lvshared, dead_tuples, nindexes); + parallel_vacuum_index(indrels, stats, lvshared, dead_tuples, nindexes, + &vacrelstats); + + /* Pop the error context stack */ + error_context_stack = errcallback.previous; vac_close_indexes(nindexes, indrels, RowExclusiveLock); table_close(onerel, ShareUpdateExclusiveLock); pfree(stats); } + +/* + * Error context callback for errors occurring during vacuum. + */ +static void +vacuum_error_callback(void *arg) +{ + LVRelStats *cbarg = arg; + + switch (cbarg->phase) + { + case VACUUM_ERRCB_PHASE_SCAN_HEAP: + if (BlockNumberIsValid(cbarg->blkno)) + errcontext("while scanning block %u of relation \"%s.%s\"", + cbarg->blkno, cbarg->relnamespace, cbarg->relname); + break; + + case VACUUM_ERRCB_PHASE_VACUUM_HEAP: + if (BlockNumberIsValid(cbarg->blkno)) + errcontext("while vacuuming block %u of relation \"%s.%s\"", + cbarg->blkno, cbarg->relnamespace, cbarg->relname); + break; + + case VACUUM_ERRCB_PHASE_VACUUM_INDEX: + errcontext("while vacuuming index \"%s\" of relation \"%s.%s\"", + cbarg->indname, cbarg->relnamespace, cbarg->relname); + break; + + case VACUUM_ERRCB_PHASE_INDEX_CLEANUP: + errcontext("while cleaning up index \"%s\" of relation \"%s.%s\"", + cbarg->indname, cbarg->relnamespace, cbarg->relname); + break; + + case VACUUM_ERRCB_PHASE_UNKNOWN: + default: + return; /* do nothing; the cbarg may not be + * initialized */ + } +} + +/* Update vacuum error callback for current phase, block and index */ +static void +update_vacuum_error_cbarg(LVRelStats *errcbarg, int phase, BlockNumber blkno, + char *indname) +{ + errcbarg->blkno = blkno; + errcbarg->phase = phase; + + /* Free index name from any previous phase */ + if (errcbarg->indname) + pfree(errcbarg->indname); + + /* For index phases, save the name of the current index for the callback */ + errcbarg->indname = indname ? pstrdup(indname) : NULL; +} diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list index fcfcf56f4f..1dee4e1ff2 100644 --- a/src/tools/pgindent/typedefs.list +++ b/src/tools/pgindent/typedefs.list @@ -582,6 +582,7 @@ EphemeralNamedRelationMetadata EphemeralNamedRelationMetadataData EquivalenceClass EquivalenceMember +errcb_phase ErrorContextCallback ErrorData EstimateDSMForeignScan_function -- 2.17.0 --rG8Oha4Ryswg1dZ8 Content-Type: text/x-diff; charset=us-ascii Content-Disposition: attachment; filename="v28-0002-Save-the-values-of-the-callback.patch" ^ permalink raw reply [nested|flat] 10+ messages in thread
* [PATCH v23 1/3] vacuum errcontext to show block being processed @ 2019-12-13 02:54 Justin Pryzby <pryzbyj@telsasoft.com> 0 siblings, 0 replies; 10+ messages in thread From: Justin Pryzby @ 2019-12-13 02:54 UTC (permalink / raw) Discussion: https://www.postgresql.org/message-id/20191120210600.GC30362@telsasoft.com --- src/backend/access/heap/vacuumlazy.c | 178 +++++++++++++++++++++++---- src/include/commands/progress.h | 1 + 2 files changed, 153 insertions(+), 26 deletions(-) diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c index 03c43efc32..f15326a24c 100644 --- a/src/backend/access/heap/vacuumlazy.c +++ b/src/backend/access/heap/vacuumlazy.c @@ -270,6 +270,8 @@ typedef struct LVParallelState typedef struct LVRelStats { + char *relnamespace; + char *relname; /* useindex = true means two-pass strategy; false means one-pass */ bool useindex; /* Overall statistics about rel */ @@ -290,8 +292,12 @@ typedef struct LVRelStats int num_index_scans; TransactionId latestRemovedXid; bool lock_waiter_detected; -} LVRelStats; + /* Used for error callback: */ + char *indname; + BlockNumber blkno; /* used only for heap operations */ + int phase; /* Reusing same constants as for progress reporting */ +} LVRelStats; /* A few variables that don't seem worth passing around as parameters */ static int elevel = -1; @@ -314,10 +320,10 @@ static void lazy_vacuum_all_indexes(Relation onerel, Relation *Irel, LVRelStats *vacrelstats, LVParallelState *lps, int nindexes); static void lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, - LVDeadTuples *dead_tuples, double reltuples); + LVDeadTuples *dead_tuples, double reltuples, LVRelStats *vacrelstats); static void lazy_cleanup_index(Relation indrel, IndexBulkDeleteResult **stats, - double reltuples, bool estimated_count); + double reltuples, bool estimated_count, LVRelStats *vacrelstats); static int lazy_vacuum_page(Relation onerel, BlockNumber blkno, Buffer buffer, int tupindex, LVRelStats *vacrelstats, Buffer *vmbuffer); static bool should_attempt_truncation(VacuumParams *params, @@ -337,13 +343,13 @@ static void lazy_parallel_vacuum_indexes(Relation *Irel, IndexBulkDeleteResult * int nindexes); static void parallel_vacuum_index(Relation *Irel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVDeadTuples *dead_tuples, - int nindexes); + int nindexes, LVRelStats *vacrelstats); static void vacuum_indexes_leader(Relation *Irel, IndexBulkDeleteResult **stats, LVRelStats *vacrelstats, LVParallelState *lps, int nindexes); static void vacuum_one_index(Relation indrel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVSharedIndStats *shared_indstats, - LVDeadTuples *dead_tuples); + LVDeadTuples *dead_tuples, LVRelStats *vacrelstats); static void lazy_cleanup_all_indexes(Relation *Irel, IndexBulkDeleteResult **stats, LVRelStats *vacrelstats, LVParallelState *lps, int nindexes); @@ -361,6 +367,9 @@ static void end_parallel_vacuum(Relation *Irel, IndexBulkDeleteResult **stats, LVParallelState *lps, int nindexes); static LVSharedIndStats *get_indstats(LVShared *lvshared, int n); static bool skip_parallel_vacuum_index(Relation indrel, LVShared *lvshared); +static void vacuum_error_callback(void *arg); +static void update_vacuum_error_cbarg(LVRelStats *errcbarg, int phase, + BlockNumber blkno, Relation rel); /* @@ -460,6 +469,9 @@ heap_vacuum_rel(Relation onerel, VacuumParams *params, vacrelstats = (LVRelStats *) palloc0(sizeof(LVRelStats)); + vacrelstats->relnamespace = get_namespace_name(RelationGetNamespace(onerel)); + vacrelstats->relname = pstrdup(RelationGetRelationName(onerel)); + vacrelstats->indname = NULL; vacrelstats->old_rel_pages = onerel->rd_rel->relpages; vacrelstats->old_live_tuples = onerel->rd_rel->reltuples; vacrelstats->num_index_scans = 0; @@ -699,7 +711,6 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, BlockNumber nblocks, blkno; HeapTupleData tuple; - char *relname; TransactionId relfrozenxid = onerel->rd_rel->relfrozenxid; TransactionId relminmxid = onerel->rd_rel->relminmxid; BlockNumber empty_pages, @@ -724,20 +735,20 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, PROGRESS_VACUUM_MAX_DEAD_TUPLES }; int64 initprog_val[3]; + ErrorContextCallback errcallback; pg_rusage_init(&ru0); - relname = RelationGetRelationName(onerel); if (aggressive) ereport(elevel, (errmsg("aggressively vacuuming \"%s.%s\"", - get_namespace_name(RelationGetNamespace(onerel)), - relname))); + vacrelstats->relnamespace, + vacrelstats->relname))); else ereport(elevel, (errmsg("vacuuming \"%s.%s\"", - get_namespace_name(RelationGetNamespace(onerel)), - relname))); + vacrelstats->relnamespace, + vacrelstats->relname))); empty_pages = vacuumed_pages = 0; next_fsm_block_to_vacuum = (BlockNumber) 0; @@ -870,6 +881,14 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, else skipping_blocks = false; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, PROGRESS_VACUUM_PHASE_SCAN_HEAP, + InvalidBlockNumber, NULL); + errcallback.callback = vacuum_error_callback; + errcallback.arg = vacrelstats; + errcallback.previous = error_context_stack; + error_context_stack = &errcallback; + for (blkno = 0; blkno < nblocks; blkno++) { Buffer buf; @@ -891,6 +910,8 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, #define FORCE_CHECK_PAGE() \ (blkno == nblocks - 1 && should_attempt_truncation(params, vacrelstats)) + vacrelstats->blkno = blkno; + pgstat_progress_update_param(PROGRESS_VACUUM_HEAP_BLKS_SCANNED, blkno); if (blkno == next_unskippable_block) @@ -1005,12 +1026,18 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, * Vacuum the Free Space Map to make newly-freed space visible on * upper-level FSM pages. Note we have not yet processed blkno. */ + update_vacuum_error_cbarg(vacrelstats, + PROGRESS_VACUUM_PHASE_VACUUM_FSM, InvalidBlockNumber, NULL); FreeSpaceMapVacuumRange(onerel, next_fsm_block_to_vacuum, blkno); next_fsm_block_to_vacuum = blkno; /* Report that we are once again scanning the heap */ pgstat_progress_update_param(PROGRESS_VACUUM_PHASE, PROGRESS_VACUUM_PHASE_SCAN_HEAP); + + /* Set the error context while continuing heap scan */ + update_vacuum_error_cbarg(vacrelstats, + PROGRESS_VACUUM_PHASE_SCAN_HEAP, blkno, NULL); } /* @@ -1534,7 +1561,7 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, && VM_ALL_VISIBLE(onerel, blkno, &vmbuffer)) { elog(WARNING, "page is not marked all-visible but visibility map bit is set in relation \"%s\" page %u", - relname, blkno); + vacrelstats->relname, blkno); visibilitymap_clear(onerel, blkno, vmbuffer, VISIBILITYMAP_VALID_BITS); } @@ -1555,7 +1582,7 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, else if (PageIsAllVisible(page) && has_dead_tuples) { elog(WARNING, "page containing dead tuples is marked as all-visible in relation \"%s\" page %u", - relname, blkno); + vacrelstats->relname, blkno); PageClearAllVisible(page); MarkBufferDirty(buf); visibilitymap_clear(onerel, blkno, vmbuffer, @@ -1651,6 +1678,9 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, if (vacrelstats->useindex) lazy_cleanup_all_indexes(Irel, indstats, vacrelstats, lps, nindexes); + /* Pop the error context stack */ + error_context_stack = errcallback.previous; + /* * End parallel mode before updating index statistics as we cannot write * during parallel mode. @@ -1744,7 +1774,7 @@ lazy_vacuum_all_indexes(Relation onerel, Relation *Irel, for (idx = 0; idx < nindexes; idx++) lazy_vacuum_index(Irel[idx], &stats[idx], vacrelstats->dead_tuples, - vacrelstats->old_live_tuples); + vacrelstats->old_live_tuples, vacrelstats); } /* Increase and report the number of index scans */ @@ -1778,8 +1808,13 @@ lazy_vacuum_heap(Relation onerel, LVRelStats *vacrelstats) PROGRESS_VACUUM_PHASE_VACUUM_HEAP); pg_rusage_init(&ru0); - npages = 0; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, + PROGRESS_VACUUM_PHASE_VACUUM_HEAP, InvalidBlockNumber, + NULL); + + npages = 0; tupindex = 0; while (tupindex < vacrelstats->dead_tuples->num_tuples) { @@ -1791,6 +1826,7 @@ lazy_vacuum_heap(Relation onerel, LVRelStats *vacrelstats) vacuum_delay_point(); tblk = ItemPointerGetBlockNumber(&vacrelstats->dead_tuples->itemptrs[tupindex]); + vacrelstats->blkno = tblk; buf = ReadBufferExtended(onerel, MAIN_FORKNUM, tblk, RBM_NORMAL, vac_strategy); if (!ConditionalLockBufferForCleanup(buf)) @@ -2083,7 +2119,7 @@ lazy_parallel_vacuum_indexes(Relation *Irel, IndexBulkDeleteResult **stats, * indexes in the case where no workers are launched. */ parallel_vacuum_index(Irel, stats, lps->lvshared, - vacrelstats->dead_tuples, nindexes); + vacrelstats->dead_tuples, nindexes, vacrelstats); /* Wait for all vacuum workers to finish */ WaitForParallelWorkersToFinish(lps->pcxt); @@ -2106,7 +2142,7 @@ lazy_parallel_vacuum_indexes(Relation *Irel, IndexBulkDeleteResult **stats, static void parallel_vacuum_index(Relation *Irel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVDeadTuples *dead_tuples, - int nindexes) + int nindexes, LVRelStats *vacrelstats) { /* * Increment the active worker count if we are able to launch any worker. @@ -2140,7 +2176,7 @@ parallel_vacuum_index(Relation *Irel, IndexBulkDeleteResult **stats, /* Do vacuum or cleanup of the index */ vacuum_one_index(Irel[idx], &(stats[idx]), lvshared, shared_indstats, - dead_tuples); + dead_tuples, vacrelstats); } /* @@ -2180,7 +2216,8 @@ vacuum_indexes_leader(Relation *Irel, IndexBulkDeleteResult **stats, if (shared_indstats == NULL || skip_parallel_vacuum_index(Irel[i], lps->lvshared)) vacuum_one_index(Irel[i], &(stats[i]), lps->lvshared, - shared_indstats, vacrelstats->dead_tuples); + shared_indstats, vacrelstats->dead_tuples, + vacrelstats); } /* @@ -2200,7 +2237,7 @@ vacuum_indexes_leader(Relation *Irel, IndexBulkDeleteResult **stats, static void vacuum_one_index(Relation indrel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVSharedIndStats *shared_indstats, - LVDeadTuples *dead_tuples) + LVDeadTuples *dead_tuples, LVRelStats *vacrelstats) { IndexBulkDeleteResult *bulkdelete_res = NULL; @@ -2220,10 +2257,10 @@ vacuum_one_index(Relation indrel, IndexBulkDeleteResult **stats, /* Do vacuum or cleanup of the index */ if (lvshared->for_cleanup) lazy_cleanup_index(indrel, stats, lvshared->reltuples, - lvshared->estimated_count); + lvshared->estimated_count, vacrelstats); else lazy_vacuum_index(indrel, stats, dead_tuples, - lvshared->reltuples); + lvshared->reltuples, vacrelstats); /* * Copy the index bulk-deletion result returned from ambulkdelete and @@ -2298,7 +2335,8 @@ lazy_cleanup_all_indexes(Relation *Irel, IndexBulkDeleteResult **stats, for (idx = 0; idx < nindexes; idx++) lazy_cleanup_index(Irel[idx], &stats[idx], vacrelstats->new_rel_tuples, - vacrelstats->tupcount_pages < vacrelstats->rel_pages); + vacrelstats->tupcount_pages < vacrelstats->rel_pages, + vacrelstats); } } @@ -2313,7 +2351,7 @@ lazy_cleanup_all_indexes(Relation *Irel, IndexBulkDeleteResult **stats, */ static void lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, - LVDeadTuples *dead_tuples, double reltuples) + LVDeadTuples *dead_tuples, double reltuples, LVRelStats *vacrelstats) { IndexVacuumInfo ivinfo; const char *msg; @@ -2329,6 +2367,11 @@ lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, ivinfo.num_heap_tuples = reltuples; ivinfo.strategy = vac_strategy; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, + PROGRESS_VACUUM_PHASE_VACUUM_INDEX, InvalidBlockNumber, + indrel); + /* Do bulk deletion */ *stats = index_bulk_delete(&ivinfo, *stats, lazy_tid_reaped, (void *) dead_tuples); @@ -2354,7 +2397,7 @@ lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, static void lazy_cleanup_index(Relation indrel, IndexBulkDeleteResult **stats, - double reltuples, bool estimated_count) + double reltuples, bool estimated_count, LVRelStats *vacrelstats) { IndexVacuumInfo ivinfo; const char *msg; @@ -2371,6 +2414,10 @@ lazy_cleanup_index(Relation indrel, ivinfo.num_heap_tuples = reltuples; ivinfo.strategy = vac_strategy; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, + PROGRESS_VACUUM_PHASE_INDEX_CLEANUP, InvalidBlockNumber, indrel); + *stats = index_vacuum_cleanup(&ivinfo, *stats); if (!(*stats)) @@ -3320,6 +3367,8 @@ parallel_vacuum_main(dsm_segment *seg, shm_toc *toc) int nindexes; char *sharedquery; IndexBulkDeleteResult **stats; + LVRelStats vacrelstats; + ErrorContextCallback errcallback; lvshared = (LVShared *) shm_toc_lookup(toc, PARALLEL_VACUUM_KEY_SHARED, false); @@ -3341,6 +3390,18 @@ parallel_vacuum_main(dsm_segment *seg, shm_toc *toc) */ onerel = table_open(lvshared->relid, ShareUpdateExclusiveLock); + /* Init vacrelstats for use as error callback by parallel worker: */ + vacrelstats.relnamespace = get_namespace_name(RelationGetNamespace(onerel)); + vacrelstats.relname = pstrdup(RelationGetRelationName(onerel)); + vacrelstats.indname = NULL; + vacrelstats.phase = -1, /* Not yet processing */ + + /* Setup error traceback support for ereport() */ + errcallback.callback = vacuum_error_callback; + errcallback.arg = &vacrelstats; + errcallback.previous = error_context_stack; + error_context_stack = &errcallback; + /* * Open all indexes. indrels are sorted in order by OID, which should be * matched to the leader's one. @@ -3370,9 +3431,74 @@ parallel_vacuum_main(dsm_segment *seg, shm_toc *toc) maintenance_work_mem = lvshared->maintenance_work_mem_worker; /* Process indexes to perform vacuum/cleanup */ - parallel_vacuum_index(indrels, stats, lvshared, dead_tuples, nindexes); + parallel_vacuum_index(indrels, stats, lvshared, dead_tuples, nindexes, &vacrelstats); vac_close_indexes(nindexes, indrels, RowExclusiveLock); table_close(onerel, ShareUpdateExclusiveLock); pfree(stats); } + +/* + * Error context callback for errors occurring during vacuum. + */ +static void +vacuum_error_callback(void *arg) +{ + LVRelStats *cbarg = arg; + + switch (cbarg->phase) { + case PROGRESS_VACUUM_PHASE_SCAN_HEAP: + if (BlockNumberIsValid(cbarg->blkno)) + errcontext("while scanning block %u of relation \"%s.%s\"", + cbarg->blkno, cbarg->relnamespace, cbarg->relname); + break; + + case PROGRESS_VACUUM_PHASE_VACUUM_HEAP: + if (BlockNumberIsValid(cbarg->blkno)) + errcontext("while vacuuming block %u of relation \"%s.%s\"", + cbarg->blkno, cbarg->relnamespace, cbarg->relname); + break; + + case PROGRESS_VACUUM_PHASE_VACUUM_INDEX: + errcontext("while vacuuming index \"%s\" of relation \"%s.%s\"", + cbarg->indname, cbarg->relnamespace, cbarg->relname); + break; + + case PROGRESS_VACUUM_PHASE_INDEX_CLEANUP: + errcontext("while cleaning up index \"%s\" of relation \"%s.%s\"", + cbarg->indname, cbarg->relnamespace, cbarg->relname); + break; + + case PROGRESS_VACUUM_PHASE_VACUUM_FSM: + errcontext("while vacuuming free space map of relation \"%s.%s\"", + cbarg->relnamespace, cbarg->relname); + break; + + case -1: /* Parallel worker not yet processing */ + case PROGRESS_VACUUM_PHASE_TRUNCATE: + case PROGRESS_VACUUM_PHASE_FINAL_CLEANUP: + default: + return; /* Shouldn't happen: do nothing */ + } +} + +/* Update vacuum error callback for current phase, block and index */ +static void +update_vacuum_error_cbarg(LVRelStats *errcbarg, int phase, BlockNumber blkno, + Relation indrel) +{ + errcbarg->blkno = blkno; + errcbarg->phase = phase; + + /* Free index name from any previous phase */ + if (errcbarg->indname) { + pfree(errcbarg->indname); + errcbarg->indname = NULL; + } + + /* For index phases, save the name of the current index for the callback */ + if (indrel) { + Assert(indrel->rd_rel->relkind == RELKIND_INDEX); + errcbarg->indname = pstrdup(RelationGetRelationName(indrel)); + } +} diff --git a/src/include/commands/progress.h b/src/include/commands/progress.h index 12e9d3d42f..5af2b8cc0c 100644 --- a/src/include/commands/progress.h +++ b/src/include/commands/progress.h @@ -33,6 +33,7 @@ #define PROGRESS_VACUUM_PHASE_INDEX_CLEANUP 4 #define PROGRESS_VACUUM_PHASE_TRUNCATE 5 #define PROGRESS_VACUUM_PHASE_FINAL_CLEANUP 6 +#define PROGRESS_VACUUM_PHASE_VACUUM_FSM 7 /* For error reporting only */ /* Progress parameters for analyze */ #define PROGRESS_ANALYZE_PHASE 0 -- 2.17.0 --HcAYCG3uE/tztfnV Content-Type: text/x-diff; charset=us-ascii Content-Disposition: attachment; filename="v23-0002-add-callback-for-truncation.patch" ^ permalink raw reply [nested|flat] 10+ messages in thread
* [PATCH v8 4/5] vacuum errcontext to show block being processed @ 2019-12-13 02:54 Justin Pryzby <pryzbyj@telsasoft.com> 0 siblings, 0 replies; 10+ messages in thread From: Justin Pryzby @ 2019-12-13 02:54 UTC (permalink / raw) As requested here. https://www.postgresql.org/message-id/20190807235154.erbmr4o4bo6vgnjv%40alap3.anarazel.de --- src/backend/access/heap/vacuumlazy.c | 37 ++++++++++++++++++++++++++++++++++++ 1 file changed, 37 insertions(+) diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c index e35fe6a..52602e2 100644 --- a/src/backend/access/heap/vacuumlazy.c +++ b/src/backend/access/heap/vacuumlazy.c @@ -138,6 +138,12 @@ typedef struct LVRelStats bool lock_waiter_detected; } LVRelStats; +typedef struct +{ + char *relname; + char *relnamespace; + BlockNumber blkno; +} vacuum_error_callback_arg; /* A few variables that don't seem worth passing around as parameters */ static int elevel = -1; @@ -178,6 +184,7 @@ static bool lazy_tid_reaped(ItemPointer itemptr, void *state); static int vac_cmp_itemptr(const void *left, const void *right); static bool heap_page_is_all_visible(Relation rel, Buffer buf, TransactionId *visibility_cutoff_xid, bool *all_frozen); +static void vacuum_error_callback(void *arg); /* @@ -609,6 +616,8 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, PROGRESS_VACUUM_MAX_DEAD_TUPLES }; int64 initprog_val[3]; + ErrorContextCallback errcallback; + vacuum_error_callback_arg errcbarg; pg_rusage_init(&ru0); @@ -650,6 +659,15 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, next_unskippable_block = 0; skipping_blocks = skip_blocks(onerel, params, &next_unskippable_block, nblocks, &vmbuffer, aggressive); + /* Setup error traceback support for ereport() */ + errcbarg.relnamespace = get_namespace_name(RelationGetNamespace(onerel)); + errcbarg.relname = relname; + errcbarg.blkno = InvalidBlockNumber; /* Not known yet */ + errcallback.callback = vacuum_error_callback; + errcallback.arg = (void *) &errcbarg; + errcallback.previous = error_context_stack; + error_context_stack = &errcallback; + for (blkno = 0; blkno < nblocks; blkno++) { Buffer buf; @@ -671,6 +689,8 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, #define FORCE_CHECK_PAGE() \ (blkno == nblocks - 1 && should_attempt_truncation(params, vacrelstats)) + errcbarg.blkno = blkno; + pgstat_progress_update_param(PROGRESS_VACUUM_HEAP_BLKS_SCANNED, blkno); if (blkno == next_unskippable_block) @@ -737,8 +757,11 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, } /* Work on all the indexes, then the heap */ + /* Don't use the errcontext handler outside this function */ + error_context_stack = errcallback.previous; lazy_vacuum_all_indexes(onerel, vacrelstats, Irel, nindexes, indstats); + error_context_stack = &errcallback; /* Remove tuples from heap */ lazy_vacuum_heap(onerel, vacrelstats); @@ -1346,6 +1369,9 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, RecordPageWithFreeSpace(onerel, blkno, freespace); } + /* Pop the error context stack */ + error_context_stack = errcallback.previous; + /* report that everything is scanned and vacuumed */ pgstat_progress_update_param(PROGRESS_VACUUM_HEAP_BLKS_SCANNED, blkno); @@ -2323,3 +2349,14 @@ heap_page_is_all_visible(Relation rel, Buffer buf, return all_visible; } + +/* + * Error context callback for errors occurring during vacuum. + */ +static void +vacuum_error_callback(void *arg) +{ + vacuum_error_callback_arg *cbarg = arg; + errcontext("while scanning block %u of relation \"%s.%s\"", + cbarg->blkno, cbarg->relnamespace, cbarg->relname); +} -- 2.7.4 --tThc/1wpZn/ma/RB Content-Type: text/x-diff; charset=us-ascii Content-Disposition: attachment; filename="v8-0005-add-errcontext-callback-in-lazy_vacuum_heap-too.patch" ^ permalink raw reply [nested|flat] 10+ messages in thread
* [PATCH v25 1/4] vacuum errcontext to show block being processed @ 2019-12-13 02:54 Justin Pryzby <pryzbyj@telsasoft.com> 0 siblings, 0 replies; 10+ messages in thread From: Justin Pryzby @ 2019-12-13 02:54 UTC (permalink / raw) Discussion: https://www.postgresql.org/message-id/20191120210600.GC30362@telsasoft.com --- src/backend/access/heap/vacuumlazy.c | 202 +++++++++++++++++++++++---- 1 file changed, 177 insertions(+), 25 deletions(-) diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c index 03c43efc32..9f0efa5aad 100644 --- a/src/backend/access/heap/vacuumlazy.c +++ b/src/backend/access/heap/vacuumlazy.c @@ -268,8 +268,19 @@ typedef struct LVParallelState int nindexes_parallel_condcleanup; } LVParallelState; +typedef enum { + VACUUM_ERRCB_PHASE_UNKNOWN, + VACUUM_ERRCB_PHASE_SCAN_HEAP, + VACUUM_ERRCB_PHASE_VACUUM_INDEX, + VACUUM_ERRCB_PHASE_VACUUM_HEAP, + VACUUM_ERRCB_PHASE_INDEX_CLEANUP, + VACUUM_ERRCB_PHASE_VACUUM_FSM, +} errcb_phase; + typedef struct LVRelStats { + char *relnamespace; + char *relname; /* useindex = true means two-pass strategy; false means one-pass */ bool useindex; /* Overall statistics about rel */ @@ -290,8 +301,12 @@ typedef struct LVRelStats int num_index_scans; TransactionId latestRemovedXid; bool lock_waiter_detected; -} LVRelStats; + /* Used for error callback: */ + char *indname; + BlockNumber blkno; /* used only for heap operations */ + errcb_phase phase; +} LVRelStats; /* A few variables that don't seem worth passing around as parameters */ static int elevel = -1; @@ -314,10 +329,10 @@ static void lazy_vacuum_all_indexes(Relation onerel, Relation *Irel, LVRelStats *vacrelstats, LVParallelState *lps, int nindexes); static void lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, - LVDeadTuples *dead_tuples, double reltuples); + LVDeadTuples *dead_tuples, double reltuples, LVRelStats *vacrelstats); static void lazy_cleanup_index(Relation indrel, IndexBulkDeleteResult **stats, - double reltuples, bool estimated_count); + double reltuples, bool estimated_count, LVRelStats *vacrelstats); static int lazy_vacuum_page(Relation onerel, BlockNumber blkno, Buffer buffer, int tupindex, LVRelStats *vacrelstats, Buffer *vmbuffer); static bool should_attempt_truncation(VacuumParams *params, @@ -337,13 +352,13 @@ static void lazy_parallel_vacuum_indexes(Relation *Irel, IndexBulkDeleteResult * int nindexes); static void parallel_vacuum_index(Relation *Irel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVDeadTuples *dead_tuples, - int nindexes); + int nindexes, LVRelStats *vacrelstats); static void vacuum_indexes_leader(Relation *Irel, IndexBulkDeleteResult **stats, LVRelStats *vacrelstats, LVParallelState *lps, int nindexes); static void vacuum_one_index(Relation indrel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVSharedIndStats *shared_indstats, - LVDeadTuples *dead_tuples); + LVDeadTuples *dead_tuples, LVRelStats *vacrelstats); static void lazy_cleanup_all_indexes(Relation *Irel, IndexBulkDeleteResult **stats, LVRelStats *vacrelstats, LVParallelState *lps, int nindexes); @@ -361,6 +376,9 @@ static void end_parallel_vacuum(Relation *Irel, IndexBulkDeleteResult **stats, LVParallelState *lps, int nindexes); static LVSharedIndStats *get_indstats(LVShared *lvshared, int n); static bool skip_parallel_vacuum_index(Relation indrel, LVShared *lvshared); +static void vacuum_error_callback(void *arg); +static void update_vacuum_error_cbarg(LVRelStats *errcbarg, int phase, + BlockNumber blkno, Relation rel); /* @@ -460,6 +478,9 @@ heap_vacuum_rel(Relation onerel, VacuumParams *params, vacrelstats = (LVRelStats *) palloc0(sizeof(LVRelStats)); + vacrelstats->relnamespace = get_namespace_name(RelationGetNamespace(onerel)); + vacrelstats->relname = pstrdup(RelationGetRelationName(onerel)); + vacrelstats->indname = NULL; vacrelstats->old_rel_pages = onerel->rd_rel->relpages; vacrelstats->old_live_tuples = onerel->rd_rel->reltuples; vacrelstats->num_index_scans = 0; @@ -699,7 +720,6 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, BlockNumber nblocks, blkno; HeapTupleData tuple; - char *relname; TransactionId relfrozenxid = onerel->rd_rel->relfrozenxid; TransactionId relminmxid = onerel->rd_rel->relminmxid; BlockNumber empty_pages, @@ -724,20 +744,20 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, PROGRESS_VACUUM_MAX_DEAD_TUPLES }; int64 initprog_val[3]; + ErrorContextCallback errcallback; pg_rusage_init(&ru0); - relname = RelationGetRelationName(onerel); if (aggressive) ereport(elevel, (errmsg("aggressively vacuuming \"%s.%s\"", - get_namespace_name(RelationGetNamespace(onerel)), - relname))); + vacrelstats->relnamespace, + vacrelstats->relname))); else ereport(elevel, (errmsg("vacuuming \"%s.%s\"", - get_namespace_name(RelationGetNamespace(onerel)), - relname))); + vacrelstats->relnamespace, + vacrelstats->relname))); empty_pages = vacuumed_pages = 0; next_fsm_block_to_vacuum = (BlockNumber) 0; @@ -870,6 +890,14 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, else skipping_blocks = false; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, VACUUM_ERRCB_PHASE_SCAN_HEAP, + InvalidBlockNumber, NULL); + errcallback.callback = vacuum_error_callback; + errcallback.arg = vacrelstats; + errcallback.previous = error_context_stack; + error_context_stack = &errcallback; + for (blkno = 0; blkno < nblocks; blkno++) { Buffer buf; @@ -891,6 +919,8 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, #define FORCE_CHECK_PAGE() \ (blkno == nblocks - 1 && should_attempt_truncation(params, vacrelstats)) + vacrelstats->blkno = blkno; + pgstat_progress_update_param(PROGRESS_VACUUM_HEAP_BLKS_SCANNED, blkno); if (blkno == next_unskippable_block) @@ -1005,12 +1035,18 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, * Vacuum the Free Space Map to make newly-freed space visible on * upper-level FSM pages. Note we have not yet processed blkno. */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_VACUUM_FSM, InvalidBlockNumber, NULL); FreeSpaceMapVacuumRange(onerel, next_fsm_block_to_vacuum, blkno); next_fsm_block_to_vacuum = blkno; /* Report that we are once again scanning the heap */ pgstat_progress_update_param(PROGRESS_VACUUM_PHASE, PROGRESS_VACUUM_PHASE_SCAN_HEAP); + + /* Set the error context while continuing heap scan */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_SCAN_HEAP, blkno, NULL); } /* @@ -1488,10 +1524,18 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, */ if (blkno - next_fsm_block_to_vacuum >= VACUUM_FSM_EVERY_PAGES) { + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_VACUUM_FSM, InvalidBlockNumber, + NULL); FreeSpaceMapVacuumRange(onerel, next_fsm_block_to_vacuum, blkno); next_fsm_block_to_vacuum = blkno; } + + /* Set the error context while continuing heap scan */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_SCAN_HEAP, blkno, NULL); + } freespace = PageGetHeapFreeSpace(page); @@ -1534,7 +1578,7 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, && VM_ALL_VISIBLE(onerel, blkno, &vmbuffer)) { elog(WARNING, "page is not marked all-visible but visibility map bit is set in relation \"%s\" page %u", - relname, blkno); + vacrelstats->relname, blkno); visibilitymap_clear(onerel, blkno, vmbuffer, VISIBILITYMAP_VALID_BITS); } @@ -1555,7 +1599,7 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, else if (PageIsAllVisible(page) && has_dead_tuples) { elog(WARNING, "page containing dead tuples is marked as all-visible in relation \"%s\" page %u", - relname, blkno); + vacrelstats->relname, blkno); PageClearAllVisible(page); MarkBufferDirty(buf); visibilitymap_clear(onerel, blkno, vmbuffer, @@ -1642,7 +1686,11 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, * not there were indexes. */ if (blkno > next_fsm_block_to_vacuum) + { + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_VACUUM_FSM, InvalidBlockNumber, NULL); FreeSpaceMapVacuumRange(onerel, next_fsm_block_to_vacuum, blkno); + } /* report all blocks vacuumed */ pgstat_progress_update_param(PROGRESS_VACUUM_HEAP_BLKS_VACUUMED, blkno); @@ -1651,6 +1699,9 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, if (vacrelstats->useindex) lazy_cleanup_all_indexes(Irel, indstats, vacrelstats, lps, nindexes); + /* Pop the error context stack */ + error_context_stack = errcallback.previous; + /* * End parallel mode before updating index statistics as we cannot write * during parallel mode. @@ -1744,7 +1795,7 @@ lazy_vacuum_all_indexes(Relation onerel, Relation *Irel, for (idx = 0; idx < nindexes; idx++) lazy_vacuum_index(Irel[idx], &stats[idx], vacrelstats->dead_tuples, - vacrelstats->old_live_tuples); + vacrelstats->old_live_tuples, vacrelstats); } /* Increase and report the number of index scans */ @@ -1777,6 +1828,10 @@ lazy_vacuum_heap(Relation onerel, LVRelStats *vacrelstats) pgstat_progress_update_param(PROGRESS_VACUUM_PHASE, PROGRESS_VACUUM_PHASE_VACUUM_HEAP); + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, VACUUM_ERRCB_PHASE_VACUUM_HEAP, + InvalidBlockNumber, NULL); + pg_rusage_init(&ru0); npages = 0; @@ -1791,6 +1846,7 @@ lazy_vacuum_heap(Relation onerel, LVRelStats *vacrelstats) vacuum_delay_point(); tblk = ItemPointerGetBlockNumber(&vacrelstats->dead_tuples->itemptrs[tupindex]); + vacrelstats->blkno = tblk; buf = ReadBufferExtended(onerel, MAIN_FORKNUM, tblk, RBM_NORMAL, vac_strategy); if (!ConditionalLockBufferForCleanup(buf)) @@ -1847,6 +1903,10 @@ lazy_vacuum_page(Relation onerel, BlockNumber blkno, Buffer buffer, pgstat_progress_update_param(PROGRESS_VACUUM_HEAP_BLKS_VACUUMED, blkno); + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, VACUUM_ERRCB_PHASE_VACUUM_HEAP, + blkno, NULL); + START_CRIT_SECTION(); for (; tupindex < dead_tuples->num_tuples; tupindex++) @@ -2083,7 +2143,7 @@ lazy_parallel_vacuum_indexes(Relation *Irel, IndexBulkDeleteResult **stats, * indexes in the case where no workers are launched. */ parallel_vacuum_index(Irel, stats, lps->lvshared, - vacrelstats->dead_tuples, nindexes); + vacrelstats->dead_tuples, nindexes, vacrelstats); /* Wait for all vacuum workers to finish */ WaitForParallelWorkersToFinish(lps->pcxt); @@ -2106,7 +2166,7 @@ lazy_parallel_vacuum_indexes(Relation *Irel, IndexBulkDeleteResult **stats, static void parallel_vacuum_index(Relation *Irel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVDeadTuples *dead_tuples, - int nindexes) + int nindexes, LVRelStats *vacrelstats) { /* * Increment the active worker count if we are able to launch any worker. @@ -2140,7 +2200,7 @@ parallel_vacuum_index(Relation *Irel, IndexBulkDeleteResult **stats, /* Do vacuum or cleanup of the index */ vacuum_one_index(Irel[idx], &(stats[idx]), lvshared, shared_indstats, - dead_tuples); + dead_tuples, vacrelstats); } /* @@ -2180,7 +2240,8 @@ vacuum_indexes_leader(Relation *Irel, IndexBulkDeleteResult **stats, if (shared_indstats == NULL || skip_parallel_vacuum_index(Irel[i], lps->lvshared)) vacuum_one_index(Irel[i], &(stats[i]), lps->lvshared, - shared_indstats, vacrelstats->dead_tuples); + shared_indstats, vacrelstats->dead_tuples, + vacrelstats); } /* @@ -2200,7 +2261,7 @@ vacuum_indexes_leader(Relation *Irel, IndexBulkDeleteResult **stats, static void vacuum_one_index(Relation indrel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVSharedIndStats *shared_indstats, - LVDeadTuples *dead_tuples) + LVDeadTuples *dead_tuples, LVRelStats *vacrelstats) { IndexBulkDeleteResult *bulkdelete_res = NULL; @@ -2220,10 +2281,10 @@ vacuum_one_index(Relation indrel, IndexBulkDeleteResult **stats, /* Do vacuum or cleanup of the index */ if (lvshared->for_cleanup) lazy_cleanup_index(indrel, stats, lvshared->reltuples, - lvshared->estimated_count); + lvshared->estimated_count, vacrelstats); else lazy_vacuum_index(indrel, stats, dead_tuples, - lvshared->reltuples); + lvshared->reltuples, vacrelstats); /* * Copy the index bulk-deletion result returned from ambulkdelete and @@ -2298,7 +2359,8 @@ lazy_cleanup_all_indexes(Relation *Irel, IndexBulkDeleteResult **stats, for (idx = 0; idx < nindexes; idx++) lazy_cleanup_index(Irel[idx], &stats[idx], vacrelstats->new_rel_tuples, - vacrelstats->tupcount_pages < vacrelstats->rel_pages); + vacrelstats->tupcount_pages < vacrelstats->rel_pages, + vacrelstats); } } @@ -2313,7 +2375,7 @@ lazy_cleanup_all_indexes(Relation *Irel, IndexBulkDeleteResult **stats, */ static void lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, - LVDeadTuples *dead_tuples, double reltuples) + LVDeadTuples *dead_tuples, double reltuples, LVRelStats *vacrelstats) { IndexVacuumInfo ivinfo; const char *msg; @@ -2329,6 +2391,11 @@ lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, ivinfo.num_heap_tuples = reltuples; ivinfo.strategy = vac_strategy; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_VACUUM_INDEX, InvalidBlockNumber, + indrel); + /* Do bulk deletion */ *stats = index_bulk_delete(&ivinfo, *stats, lazy_tid_reaped, (void *) dead_tuples); @@ -2354,7 +2421,7 @@ lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, static void lazy_cleanup_index(Relation indrel, IndexBulkDeleteResult **stats, - double reltuples, bool estimated_count) + double reltuples, bool estimated_count, LVRelStats *vacrelstats) { IndexVacuumInfo ivinfo; const char *msg; @@ -2371,6 +2438,10 @@ lazy_cleanup_index(Relation indrel, ivinfo.num_heap_tuples = reltuples; ivinfo.strategy = vac_strategy; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_INDEX_CLEANUP, InvalidBlockNumber, indrel); + *stats = index_vacuum_cleanup(&ivinfo, *stats); if (!(*stats)) @@ -3320,6 +3391,8 @@ parallel_vacuum_main(dsm_segment *seg, shm_toc *toc) int nindexes; char *sharedquery; IndexBulkDeleteResult **stats; + LVRelStats vacrelstats; + ErrorContextCallback errcallback; lvshared = (LVShared *) shm_toc_lookup(toc, PARALLEL_VACUUM_KEY_SHARED, false); @@ -3369,10 +3442,89 @@ parallel_vacuum_main(dsm_segment *seg, shm_toc *toc) if (lvshared->maintenance_work_mem_worker > 0) maintenance_work_mem = lvshared->maintenance_work_mem_worker; + /* Init vacrelstats for use as error callback arg by parallel worker */ + vacrelstats.relnamespace = get_namespace_name(RelationGetNamespace(onerel)); + vacrelstats.relname = pstrdup(RelationGetRelationName(onerel)); + vacrelstats.indname = NULL; + vacrelstats.phase = VACUUM_ERRCB_PHASE_UNKNOWN; /* Not yet processing */ + + /* Setup error traceback support for ereport() */ + errcallback.callback = vacuum_error_callback; + errcallback.arg = &vacrelstats; + errcallback.previous = error_context_stack; + error_context_stack = &errcallback; + /* Process indexes to perform vacuum/cleanup */ - parallel_vacuum_index(indrels, stats, lvshared, dead_tuples, nindexes); + parallel_vacuum_index(indrels, stats, lvshared, dead_tuples, nindexes, + &vacrelstats); + + /* Pop the error context stack */ + error_context_stack = errcallback.previous; vac_close_indexes(nindexes, indrels, RowExclusiveLock); table_close(onerel, ShareUpdateExclusiveLock); pfree(stats); } + +/* + * Error context callback for errors occurring during vacuum. + */ +static void +vacuum_error_callback(void *arg) +{ + LVRelStats *cbarg = arg; + + switch (cbarg->phase) { + case VACUUM_ERRCB_PHASE_SCAN_HEAP: + if (BlockNumberIsValid(cbarg->blkno)) + errcontext("while scanning block %u of relation \"%s.%s\"", + cbarg->blkno, cbarg->relnamespace, cbarg->relname); + break; + + case VACUUM_ERRCB_PHASE_VACUUM_HEAP: + if (BlockNumberIsValid(cbarg->blkno)) + errcontext("while vacuuming block %u of relation \"%s.%s\"", + cbarg->blkno, cbarg->relnamespace, cbarg->relname); + break; + + case VACUUM_ERRCB_PHASE_VACUUM_INDEX: + errcontext("while vacuuming index \"%s\" of relation \"%s.%s\"", + cbarg->indname, cbarg->relnamespace, cbarg->relname); + break; + + case VACUUM_ERRCB_PHASE_INDEX_CLEANUP: + errcontext("while cleaning up index \"%s\" of relation \"%s.%s\"", + cbarg->indname, cbarg->relnamespace, cbarg->relname); + break; + + case VACUUM_ERRCB_PHASE_VACUUM_FSM: + errcontext("while vacuuming free space map of relation \"%s.%s\"", + cbarg->relnamespace, cbarg->relname); + break; + + case VACUUM_ERRCB_PHASE_UNKNOWN: + default: + return; /* do nothing; the cbarg maybe isn't yet initialized */ + } +} + +/* Update vacuum error callback for current phase, block and index */ +static void +update_vacuum_error_cbarg(LVRelStats *errcbarg, int phase, BlockNumber blkno, + Relation indrel) +{ + errcbarg->blkno = blkno; + errcbarg->phase = phase; + + /* Free index name from any previous phase */ + if (errcbarg->indname) { + pfree(errcbarg->indname); + errcbarg->indname = NULL; + } + + /* For index phases, save the name of the current index for the callback */ + if (indrel) { + Assert(indrel->rd_rel->relkind == RELKIND_INDEX); + errcbarg->indname = pstrdup(RelationGetRelationName(indrel)); + } +} -- 2.17.0 --EcLUEBHJhOmOUQ0C Content-Type: text/x-diff; charset=us-ascii Content-Disposition: attachment; filename="v25-0002-Drop-reltuples.patch" ^ permalink raw reply [nested|flat] 10+ messages in thread
* [PATCH v24 1/4] vacuum errcontext to show block being processed @ 2019-12-13 02:54 Justin Pryzby <pryzbyj@telsasoft.com> 0 siblings, 0 replies; 10+ messages in thread From: Justin Pryzby @ 2019-12-13 02:54 UTC (permalink / raw) Discussion: https://www.postgresql.org/message-id/20191120210600.GC30362@telsasoft.com --- src/backend/access/heap/vacuumlazy.c | 192 +++++++++++++++++++++++---- 1 file changed, 167 insertions(+), 25 deletions(-) diff --git a/src/backend/access/heap/vacuumlazy.c b/src/backend/access/heap/vacuumlazy.c index 03c43efc32..084b43c178 100644 --- a/src/backend/access/heap/vacuumlazy.c +++ b/src/backend/access/heap/vacuumlazy.c @@ -268,8 +268,19 @@ typedef struct LVParallelState int nindexes_parallel_condcleanup; } LVParallelState; +typedef enum { + VACUUM_ERRCB_PHASE_UNKNOWN, + VACUUM_ERRCB_PHASE_SCAN_HEAP, + VACUUM_ERRCB_PHASE_VACUUM_INDEX, + VACUUM_ERRCB_PHASE_VACUUM_HEAP, + VACUUM_ERRCB_PHASE_INDEX_CLEANUP, + VACUUM_ERRCB_PHASE_VACUUM_FSM, +} errcb_phase; + typedef struct LVRelStats { + char *relnamespace; + char *relname; /* useindex = true means two-pass strategy; false means one-pass */ bool useindex; /* Overall statistics about rel */ @@ -290,8 +301,12 @@ typedef struct LVRelStats int num_index_scans; TransactionId latestRemovedXid; bool lock_waiter_detected; -} LVRelStats; + /* Used for error callback: */ + char *indname; + BlockNumber blkno; /* used only for heap operations */ + errcb_phase phase; +} LVRelStats; /* A few variables that don't seem worth passing around as parameters */ static int elevel = -1; @@ -314,10 +329,10 @@ static void lazy_vacuum_all_indexes(Relation onerel, Relation *Irel, LVRelStats *vacrelstats, LVParallelState *lps, int nindexes); static void lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, - LVDeadTuples *dead_tuples, double reltuples); + LVDeadTuples *dead_tuples, double reltuples, LVRelStats *vacrelstats); static void lazy_cleanup_index(Relation indrel, IndexBulkDeleteResult **stats, - double reltuples, bool estimated_count); + double reltuples, bool estimated_count, LVRelStats *vacrelstats); static int lazy_vacuum_page(Relation onerel, BlockNumber blkno, Buffer buffer, int tupindex, LVRelStats *vacrelstats, Buffer *vmbuffer); static bool should_attempt_truncation(VacuumParams *params, @@ -337,13 +352,13 @@ static void lazy_parallel_vacuum_indexes(Relation *Irel, IndexBulkDeleteResult * int nindexes); static void parallel_vacuum_index(Relation *Irel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVDeadTuples *dead_tuples, - int nindexes); + int nindexes, LVRelStats *vacrelstats); static void vacuum_indexes_leader(Relation *Irel, IndexBulkDeleteResult **stats, LVRelStats *vacrelstats, LVParallelState *lps, int nindexes); static void vacuum_one_index(Relation indrel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVSharedIndStats *shared_indstats, - LVDeadTuples *dead_tuples); + LVDeadTuples *dead_tuples, LVRelStats *vacrelstats); static void lazy_cleanup_all_indexes(Relation *Irel, IndexBulkDeleteResult **stats, LVRelStats *vacrelstats, LVParallelState *lps, int nindexes); @@ -361,6 +376,9 @@ static void end_parallel_vacuum(Relation *Irel, IndexBulkDeleteResult **stats, LVParallelState *lps, int nindexes); static LVSharedIndStats *get_indstats(LVShared *lvshared, int n); static bool skip_parallel_vacuum_index(Relation indrel, LVShared *lvshared); +static void vacuum_error_callback(void *arg); +static void update_vacuum_error_cbarg(LVRelStats *errcbarg, int phase, + BlockNumber blkno, Relation rel); /* @@ -460,6 +478,9 @@ heap_vacuum_rel(Relation onerel, VacuumParams *params, vacrelstats = (LVRelStats *) palloc0(sizeof(LVRelStats)); + vacrelstats->relnamespace = get_namespace_name(RelationGetNamespace(onerel)); + vacrelstats->relname = pstrdup(RelationGetRelationName(onerel)); + vacrelstats->indname = NULL; vacrelstats->old_rel_pages = onerel->rd_rel->relpages; vacrelstats->old_live_tuples = onerel->rd_rel->reltuples; vacrelstats->num_index_scans = 0; @@ -699,7 +720,6 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, BlockNumber nblocks, blkno; HeapTupleData tuple; - char *relname; TransactionId relfrozenxid = onerel->rd_rel->relfrozenxid; TransactionId relminmxid = onerel->rd_rel->relminmxid; BlockNumber empty_pages, @@ -724,20 +744,20 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, PROGRESS_VACUUM_MAX_DEAD_TUPLES }; int64 initprog_val[3]; + ErrorContextCallback errcallback; pg_rusage_init(&ru0); - relname = RelationGetRelationName(onerel); if (aggressive) ereport(elevel, (errmsg("aggressively vacuuming \"%s.%s\"", - get_namespace_name(RelationGetNamespace(onerel)), - relname))); + vacrelstats->relnamespace, + vacrelstats->relname))); else ereport(elevel, (errmsg("vacuuming \"%s.%s\"", - get_namespace_name(RelationGetNamespace(onerel)), - relname))); + vacrelstats->relnamespace, + vacrelstats->relname))); empty_pages = vacuumed_pages = 0; next_fsm_block_to_vacuum = (BlockNumber) 0; @@ -870,6 +890,14 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, else skipping_blocks = false; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, VACUUM_ERRCB_PHASE_SCAN_HEAP, + InvalidBlockNumber, NULL); + errcallback.callback = vacuum_error_callback; + errcallback.arg = vacrelstats; + errcallback.previous = error_context_stack; + error_context_stack = &errcallback; + for (blkno = 0; blkno < nblocks; blkno++) { Buffer buf; @@ -891,6 +919,8 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, #define FORCE_CHECK_PAGE() \ (blkno == nblocks - 1 && should_attempt_truncation(params, vacrelstats)) + vacrelstats->blkno = blkno; + pgstat_progress_update_param(PROGRESS_VACUUM_HEAP_BLKS_SCANNED, blkno); if (blkno == next_unskippable_block) @@ -1005,12 +1035,18 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, * Vacuum the Free Space Map to make newly-freed space visible on * upper-level FSM pages. Note we have not yet processed blkno. */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_VACUUM_FSM, InvalidBlockNumber, NULL); FreeSpaceMapVacuumRange(onerel, next_fsm_block_to_vacuum, blkno); next_fsm_block_to_vacuum = blkno; /* Report that we are once again scanning the heap */ pgstat_progress_update_param(PROGRESS_VACUUM_PHASE, PROGRESS_VACUUM_PHASE_SCAN_HEAP); + + /* Set the error context while continuing heap scan */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_SCAN_HEAP, blkno, NULL); } /* @@ -1488,9 +1524,15 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, */ if (blkno - next_fsm_block_to_vacuum >= VACUUM_FSM_EVERY_PAGES) { + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_VACUUM_FSM, InvalidBlockNumber, + NULL); FreeSpaceMapVacuumRange(onerel, next_fsm_block_to_vacuum, blkno); next_fsm_block_to_vacuum = blkno; + /* Set the error context while continuing heap scan */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_SCAN_HEAP, blkno, NULL); } } @@ -1534,7 +1576,7 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, && VM_ALL_VISIBLE(onerel, blkno, &vmbuffer)) { elog(WARNING, "page is not marked all-visible but visibility map bit is set in relation \"%s\" page %u", - relname, blkno); + vacrelstats->relname, blkno); visibilitymap_clear(onerel, blkno, vmbuffer, VISIBILITYMAP_VALID_BITS); } @@ -1555,7 +1597,7 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, else if (PageIsAllVisible(page) && has_dead_tuples) { elog(WARNING, "page containing dead tuples is marked as all-visible in relation \"%s\" page %u", - relname, blkno); + vacrelstats->relname, blkno); PageClearAllVisible(page); MarkBufferDirty(buf); visibilitymap_clear(onerel, blkno, vmbuffer, @@ -1642,7 +1684,11 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, * not there were indexes. */ if (blkno > next_fsm_block_to_vacuum) + { + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_VACUUM_FSM, InvalidBlockNumber, NULL); FreeSpaceMapVacuumRange(onerel, next_fsm_block_to_vacuum, blkno); + } /* report all blocks vacuumed */ pgstat_progress_update_param(PROGRESS_VACUUM_HEAP_BLKS_VACUUMED, blkno); @@ -1651,6 +1697,9 @@ lazy_scan_heap(Relation onerel, VacuumParams *params, LVRelStats *vacrelstats, if (vacrelstats->useindex) lazy_cleanup_all_indexes(Irel, indstats, vacrelstats, lps, nindexes); + /* Pop the error context stack */ + error_context_stack = errcallback.previous; + /* * End parallel mode before updating index statistics as we cannot write * during parallel mode. @@ -1744,7 +1793,7 @@ lazy_vacuum_all_indexes(Relation onerel, Relation *Irel, for (idx = 0; idx < nindexes; idx++) lazy_vacuum_index(Irel[idx], &stats[idx], vacrelstats->dead_tuples, - vacrelstats->old_live_tuples); + vacrelstats->old_live_tuples, vacrelstats); } /* Increase and report the number of index scans */ @@ -1847,6 +1896,10 @@ lazy_vacuum_page(Relation onerel, BlockNumber blkno, Buffer buffer, pgstat_progress_update_param(PROGRESS_VACUUM_HEAP_BLKS_VACUUMED, blkno); + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, VACUUM_ERRCB_PHASE_VACUUM_HEAP, + blkno, NULL); + START_CRIT_SECTION(); for (; tupindex < dead_tuples->num_tuples; tupindex++) @@ -2083,7 +2136,7 @@ lazy_parallel_vacuum_indexes(Relation *Irel, IndexBulkDeleteResult **stats, * indexes in the case where no workers are launched. */ parallel_vacuum_index(Irel, stats, lps->lvshared, - vacrelstats->dead_tuples, nindexes); + vacrelstats->dead_tuples, nindexes, vacrelstats); /* Wait for all vacuum workers to finish */ WaitForParallelWorkersToFinish(lps->pcxt); @@ -2106,7 +2159,7 @@ lazy_parallel_vacuum_indexes(Relation *Irel, IndexBulkDeleteResult **stats, static void parallel_vacuum_index(Relation *Irel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVDeadTuples *dead_tuples, - int nindexes) + int nindexes, LVRelStats *vacrelstats) { /* * Increment the active worker count if we are able to launch any worker. @@ -2140,7 +2193,7 @@ parallel_vacuum_index(Relation *Irel, IndexBulkDeleteResult **stats, /* Do vacuum or cleanup of the index */ vacuum_one_index(Irel[idx], &(stats[idx]), lvshared, shared_indstats, - dead_tuples); + dead_tuples, vacrelstats); } /* @@ -2180,7 +2233,8 @@ vacuum_indexes_leader(Relation *Irel, IndexBulkDeleteResult **stats, if (shared_indstats == NULL || skip_parallel_vacuum_index(Irel[i], lps->lvshared)) vacuum_one_index(Irel[i], &(stats[i]), lps->lvshared, - shared_indstats, vacrelstats->dead_tuples); + shared_indstats, vacrelstats->dead_tuples, + vacrelstats); } /* @@ -2200,7 +2254,7 @@ vacuum_indexes_leader(Relation *Irel, IndexBulkDeleteResult **stats, static void vacuum_one_index(Relation indrel, IndexBulkDeleteResult **stats, LVShared *lvshared, LVSharedIndStats *shared_indstats, - LVDeadTuples *dead_tuples) + LVDeadTuples *dead_tuples, LVRelStats *vacrelstats) { IndexBulkDeleteResult *bulkdelete_res = NULL; @@ -2220,10 +2274,10 @@ vacuum_one_index(Relation indrel, IndexBulkDeleteResult **stats, /* Do vacuum or cleanup of the index */ if (lvshared->for_cleanup) lazy_cleanup_index(indrel, stats, lvshared->reltuples, - lvshared->estimated_count); + lvshared->estimated_count, vacrelstats); else lazy_vacuum_index(indrel, stats, dead_tuples, - lvshared->reltuples); + lvshared->reltuples, vacrelstats); /* * Copy the index bulk-deletion result returned from ambulkdelete and @@ -2298,7 +2352,8 @@ lazy_cleanup_all_indexes(Relation *Irel, IndexBulkDeleteResult **stats, for (idx = 0; idx < nindexes; idx++) lazy_cleanup_index(Irel[idx], &stats[idx], vacrelstats->new_rel_tuples, - vacrelstats->tupcount_pages < vacrelstats->rel_pages); + vacrelstats->tupcount_pages < vacrelstats->rel_pages, + vacrelstats); } } @@ -2313,7 +2368,7 @@ lazy_cleanup_all_indexes(Relation *Irel, IndexBulkDeleteResult **stats, */ static void lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, - LVDeadTuples *dead_tuples, double reltuples) + LVDeadTuples *dead_tuples, double reltuples, LVRelStats *vacrelstats) { IndexVacuumInfo ivinfo; const char *msg; @@ -2329,6 +2384,11 @@ lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, ivinfo.num_heap_tuples = reltuples; ivinfo.strategy = vac_strategy; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_VACUUM_INDEX, InvalidBlockNumber, + indrel); + /* Do bulk deletion */ *stats = index_bulk_delete(&ivinfo, *stats, lazy_tid_reaped, (void *) dead_tuples); @@ -2354,7 +2414,7 @@ lazy_vacuum_index(Relation indrel, IndexBulkDeleteResult **stats, static void lazy_cleanup_index(Relation indrel, IndexBulkDeleteResult **stats, - double reltuples, bool estimated_count) + double reltuples, bool estimated_count, LVRelStats *vacrelstats) { IndexVacuumInfo ivinfo; const char *msg; @@ -2371,6 +2431,10 @@ lazy_cleanup_index(Relation indrel, ivinfo.num_heap_tuples = reltuples; ivinfo.strategy = vac_strategy; + /* Setup error traceback support for ereport() */ + update_vacuum_error_cbarg(vacrelstats, + VACUUM_ERRCB_PHASE_INDEX_CLEANUP, InvalidBlockNumber, indrel); + *stats = index_vacuum_cleanup(&ivinfo, *stats); if (!(*stats)) @@ -3320,6 +3384,8 @@ parallel_vacuum_main(dsm_segment *seg, shm_toc *toc) int nindexes; char *sharedquery; IndexBulkDeleteResult **stats; + LVRelStats vacrelstats; + ErrorContextCallback errcallback; lvshared = (LVShared *) shm_toc_lookup(toc, PARALLEL_VACUUM_KEY_SHARED, false); @@ -3341,6 +3407,18 @@ parallel_vacuum_main(dsm_segment *seg, shm_toc *toc) */ onerel = table_open(lvshared->relid, ShareUpdateExclusiveLock); + /* Init vacrelstats for use as error callback by parallel worker: */ + vacrelstats.relnamespace = get_namespace_name(RelationGetNamespace(onerel)); + vacrelstats.relname = pstrdup(RelationGetRelationName(onerel)); + vacrelstats.indname = NULL; + vacrelstats.phase = VACUUM_ERRCB_PHASE_UNKNOWN; /* Not yet processing */ + + /* Setup error traceback support for ereport() */ + errcallback.callback = vacuum_error_callback; + errcallback.arg = &vacrelstats; + errcallback.previous = error_context_stack; + error_context_stack = &errcallback; + /* * Open all indexes. indrels are sorted in order by OID, which should be * matched to the leader's one. @@ -3370,9 +3448,73 @@ parallel_vacuum_main(dsm_segment *seg, shm_toc *toc) maintenance_work_mem = lvshared->maintenance_work_mem_worker; /* Process indexes to perform vacuum/cleanup */ - parallel_vacuum_index(indrels, stats, lvshared, dead_tuples, nindexes); + parallel_vacuum_index(indrels, stats, lvshared, dead_tuples, nindexes, + &vacrelstats); vac_close_indexes(nindexes, indrels, RowExclusiveLock); table_close(onerel, ShareUpdateExclusiveLock); pfree(stats); } + +/* + * Error context callback for errors occurring during vacuum. + */ +static void +vacuum_error_callback(void *arg) +{ + LVRelStats *cbarg = arg; + + switch (cbarg->phase) { + case VACUUM_ERRCB_PHASE_SCAN_HEAP: + if (BlockNumberIsValid(cbarg->blkno)) + errcontext("while scanning block %u of relation \"%s.%s\"", + cbarg->blkno, cbarg->relnamespace, cbarg->relname); + break; + + case VACUUM_ERRCB_PHASE_VACUUM_HEAP: + if (BlockNumberIsValid(cbarg->blkno)) + errcontext("while vacuuming block %u of relation \"%s.%s\"", + cbarg->blkno, cbarg->relnamespace, cbarg->relname); + break; + + case VACUUM_ERRCB_PHASE_VACUUM_INDEX: + errcontext("while vacuuming index \"%s\" of relation \"%s.%s\"", + cbarg->indname, cbarg->relnamespace, cbarg->relname); + break; + + case VACUUM_ERRCB_PHASE_INDEX_CLEANUP: + errcontext("while cleaning up index \"%s\" of relation \"%s.%s\"", + cbarg->indname, cbarg->relnamespace, cbarg->relname); + break; + + case VACUUM_ERRCB_PHASE_VACUUM_FSM: + errcontext("while vacuuming free space map of relation \"%s.%s\"", + cbarg->relnamespace, cbarg->relname); + break; + + case VACUUM_ERRCB_PHASE_UNKNOWN: + default: + return; /* do nothing; the cbarg maybe isn't yet initialized */ + } +} + +/* Update vacuum error callback for current phase, block and index */ +static void +update_vacuum_error_cbarg(LVRelStats *errcbarg, int phase, BlockNumber blkno, + Relation indrel) +{ + errcbarg->blkno = blkno; + errcbarg->phase = phase; + + /* Free index name from any previous phase */ + if (errcbarg->indname) { + pfree(errcbarg->indname); + errcbarg->indname = NULL; + } + + /* For index phases, save the name of the current index for the callback */ + if (indrel) { + Assert(indrel->rd_rel->relkind == RELKIND_INDEX); + errcbarg->indname = pstrdup(RelationGetRelationName(indrel)); + } +} -- 2.17.0 --a8Wt8u1KmwUX3Y2C Content-Type: text/x-diff; charset=us-ascii Content-Disposition: attachment; filename="v24-0002-Drop-reltuples.patch" ^ permalink raw reply [nested|flat] 10+ messages in thread
* [PATCH 1/2] Error out any process that would block at REPACK @ 2026-04-01 15:35 Antonin Houska <ah@cybertec.at> 0 siblings, 0 replies; 10+ messages in thread From: Antonin Houska @ 2026-04-01 15:35 UTC (permalink / raw) Any process waiting on REPACK to release its lock would actually cause it to deadlock when it tries to upgrade its lock to AEL, losing all work done to that point. We avoid this by teaching the deadlock detector to raise an error when this condition is detected. --- src/backend/commands/repack.c | 47 ++++++++--- src/backend/storage/lmgr/deadlock.c | 15 ++++ src/include/storage/proc.h | 6 +- src/test/modules/injection_points/Makefile | 1 + .../expected/repack_deadlock.out | 63 ++++++++++++++ src/test/modules/injection_points/meson.build | 1 + .../specs/repack_deadlock.spec | 83 +++++++++++++++++++ 7 files changed, 201 insertions(+), 15 deletions(-) create mode 100644 src/test/modules/injection_points/expected/repack_deadlock.out create mode 100644 src/test/modules/injection_points/specs/repack_deadlock.spec diff --git a/src/backend/commands/repack.c b/src/backend/commands/repack.c index 20dad22c4b7..a5f5df77291 100644 --- a/src/backend/commands/repack.c +++ b/src/backend/commands/repack.c @@ -285,6 +285,18 @@ ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel) * to understand and we don't lose any functionality. */ PreventInTransactionBlock(isTopLevel, "REPACK (CONCURRENTLY)"); + + /* + * Also set the PROC_IN_CONCURRENT_REPACK flag. This makes the + * deadlock checker cause anyone that would conflict with us to error + * out. It's important to set this flag ahead of actually locking the + * relation; it won't of course affect anyone until we do have a lock + * that others can conflict with. + */ + LWLockAcquire(ProcArrayLock, LW_EXCLUSIVE); + MyProc->statusFlags |= PROC_IN_CONCURRENT_REPACK; + ProcGlobal->statusFlags[MyProc->pgxactoff] = MyProc->statusFlags; + LWLockRelease(ProcArrayLock); } /* @@ -489,11 +501,8 @@ RepackLockLevel(bool concurrent) * If indexOid is InvalidOid, the table will be rewritten in physical order * instead of index order. * - * Note that, in the concurrent case, the function releases the lock at some - * point, in order to get AccessExclusiveLock for the final steps (i.e. to - * swap the relation files). To make things simpler, the caller should expect - * OldHeap to be closed on return, regardless CLUOPT_CONCURRENT. (The - * AccessExclusiveLock is kept till the end of the transaction.) + * On return, OldHeap is closed but locked with AccessExclusiveLock - the lock + * will be released at end of the transaction. * * 'cmd' indicates which command is being executed, to be used for error * messages. @@ -1002,10 +1011,8 @@ rebuild_relation(Relation OldHeap, Relation index, bool verbose, * Note that the worker has to wait for all transactions with XID * already assigned to finish. If some of those transactions is * waiting for a lock conflicting with ShareUpdateExclusiveLock on our - * table (e.g. it runs CREATE INDEX), we can end up in a deadlock. - * Not sure this risk is worth unlocking/locking the table (and its - * clustering index) and checking again if it's still eligible for - * REPACK CONCURRENTLY. + * table (e.g. it runs CREATE INDEX), it should encounter ERROR in the + * deadlock checking code. */ start_repack_decoding_worker(tableOid); @@ -3090,7 +3097,19 @@ rebuild_relation_finish_concurrent(Relation NewHeap, Relation OldHeap, LockRelationOid(OldHeap->rd_rel->reltoastrelid, AccessExclusiveLock); /* - * Tuples and pages of the old heap will be gone, but the heap will stay. + * Now that we have all access-exclusive locks on all relations, we no + * longer want other processes to error out when trying to acquire a + * conflicting lock. Therefore, unset our flag. + */ + LWLockAcquire(ProcArrayLock, LW_EXCLUSIVE); + MyProc->statusFlags &= ~PROC_IN_CONCURRENT_REPACK; + ProcGlobal->statusFlags[MyProc->pgxactoff] = MyProc->statusFlags; + LWLockRelease(ProcArrayLock); + + /* + * Tuples and pages of the old heap will be gone, but the heap itself will + * stay. In order for predicate locks to continue to work, convert them + * to relation-level locks. We do this both for table and indexes. */ TransferPredicateLocksToHeapRelation(OldHeap); foreach_ptr(RelationData, index, indexrels) @@ -3364,9 +3383,11 @@ start_repack_decoding_worker(Oid relid) /* * The decoding setup must be done before the caller can have XID assigned - * for any reason, otherwise the worker might end up in a deadlock, - * waiting for the caller's transaction to end. Therefore wait here until - * the worker indicates that it has the logical decoding initialized. + * for any reason, otherwise the worker might end up waiting for the + * caller's transaction to end. (Deadlock detector does not consider this + * a conflict because the worker is in the same locking group as the + * backend that launched it.) Therefore wait here until the worker + * indicates that it has the logical decoding initialized. */ ConditionVariablePrepareToSleep(&shared->cv); for (;;) diff --git a/src/backend/storage/lmgr/deadlock.c b/src/backend/storage/lmgr/deadlock.c index b8962d875b6..c20ac682b0d 100644 --- a/src/backend/storage/lmgr/deadlock.c +++ b/src/backend/storage/lmgr/deadlock.c @@ -620,6 +620,21 @@ FindLockCycleRecurseMember(PGPROC *checkProc, proc->statusFlags & PROC_IS_AUTOVACUUM) blocking_autovacuum_proc = proc; + /* + * Similarly, if we note that we're blocked by some + * process running REPACK (CONCURRENTLY), just fail. That + * process is going to upgrade its lock at some point, and + * it would be inappropriate for any other process to + * cause that to fail. + */ + if (checkProc == MyProc && + proc->statusFlags & PROC_IN_CONCURRENT_REPACK) + ereport(ERROR, + errcode(ERRCODE_OBJECT_IN_USE), + errmsg("could not wait for concurrent REPACK"), + errdetail("Process %d waits for REPACK running on process %d", + MyProc->pid, proc->pid)); + /* We're done looking at this proclock */ break; } diff --git a/src/include/storage/proc.h b/src/include/storage/proc.h index 3e1d1fad5f9..76c6bb44251 100644 --- a/src/include/storage/proc.h +++ b/src/include/storage/proc.h @@ -70,10 +70,12 @@ struct XidCache #define PROC_AFFECTS_ALL_HORIZONS 0x20 /* this proc's xmin must be * included in vacuum horizons * in all databases */ +#define PROC_IN_CONCURRENT_REPACK 0x40 /* REPACK (CONCURRENTLY) */ -/* flags reset at EOXact */ +/* flags reset at EOXact. A bit of a misnomer ... */ #define PROC_VACUUM_STATE_MASK \ - (PROC_IN_VACUUM | PROC_IN_SAFE_IC | PROC_VACUUM_FOR_WRAPAROUND) + (PROC_IN_VACUUM | PROC_IN_SAFE_IC | PROC_VACUUM_FOR_WRAPAROUND | \ + PROC_IN_CONCURRENT_REPACK) /* * Xmin-related flags. Make sure any flags that affect how the process' Xmin diff --git a/src/test/modules/injection_points/Makefile b/src/test/modules/injection_points/Makefile index f057d143d1a..13c873969d1 100644 --- a/src/test/modules/injection_points/Makefile +++ b/src/test/modules/injection_points/Makefile @@ -15,6 +15,7 @@ REGRESS_OPTS = --dlpath=$(top_builddir)/src/test/regress ISOLATION = basic \ inplace \ repack \ + repack_deadlock \ repack_toast \ syscache-update-pruned \ heap_lock_update diff --git a/src/test/modules/injection_points/expected/repack_deadlock.out b/src/test/modules/injection_points/expected/repack_deadlock.out new file mode 100644 index 00000000000..a86e4767536 --- /dev/null +++ b/src/test/modules/injection_points/expected/repack_deadlock.out @@ -0,0 +1,63 @@ +Parsed test spec with 2 sessions + +starting permutation: wait_before_lock add_column wakeup_before_lock check1 +injection_points_attach +----------------------- + +(1 row) + +step wait_before_lock: + REPACK (CONCURRENTLY) repack_deadlock USING INDEX repack_deadlock_pkey; + <waiting ...> +step add_column: + alter table repack_deadlock add column noise text; + <waiting ...> +step add_column: <... completed> +ERROR: could not wait for concurrent REPACK +step wakeup_before_lock: + SELECT injection_points_wakeup('repack-concurrently-before-lock'); + +injection_points_wakeup +----------------------- + +(1 row) + +step wait_before_lock: <... completed> +step check1: + INSERT INTO relfilenodes(node) + SELECT relfilenode FROM pg_class WHERE relname='repack_deadlock'; + + SELECT count(DISTINCT node) FROM relfilenodes; + + SELECT i, j FROM repack_deadlock ORDER BY i, j; + + INSERT INTO data_s1(i, j) + SELECT i, j FROM repack_deadlock; + + SELECT count(*) + FROM data_s1 d1 FULL JOIN data_s2 d2 USING (i, j) + WHERE d1.i ISNULL OR d2.i ISNULL; + +count +----- + 1 +(1 row) + +i|j +-+- +1|1 +2|2 +3|3 +4|4 +(4 rows) + +count +----- + 4 +(1 row) + +injection_points_detach +----------------------- + +(1 row) + diff --git a/src/test/modules/injection_points/meson.build b/src/test/modules/injection_points/meson.build index fb1418e2caa..ead18818c83 100644 --- a/src/test/modules/injection_points/meson.build +++ b/src/test/modules/injection_points/meson.build @@ -46,6 +46,7 @@ tests += { 'basic', 'inplace', 'repack', + 'repack_deadlock', 'repack_toast', 'syscache-update-pruned', 'heap_lock_update', diff --git a/src/test/modules/injection_points/specs/repack_deadlock.spec b/src/test/modules/injection_points/specs/repack_deadlock.spec new file mode 100644 index 00000000000..9d23a6588c2 --- /dev/null +++ b/src/test/modules/injection_points/specs/repack_deadlock.spec @@ -0,0 +1,83 @@ +# Test REPACK with a concurrent transaction that would cause a deadlock +setup +{ + CREATE EXTENSION injection_points; + + CREATE TABLE repack_deadlock(i int PRIMARY KEY, j int); + INSERT INTO repack_deadlock(i, j) VALUES (1, 1), (2, 2), (3, 3), (4, 4); + + CREATE TABLE relfilenodes(node oid); + + CREATE TABLE data_s1(i int, j int); + CREATE TABLE data_s2(i int, j int); +} + +teardown +{ + DROP TABLE repack_deadlock; + DROP EXTENSION injection_points; + + DROP TABLE relfilenodes; + DROP TABLE data_s1; + DROP TABLE data_s2; +} + +session s1 +setup +{ + SELECT injection_points_set_local(); + SELECT injection_points_attach('repack-concurrently-before-lock', 'wait'); +} +# Perform the initial load and wait for s2 to do some data changes. +step wait_before_lock +{ + REPACK (CONCURRENTLY) repack_deadlock USING INDEX repack_deadlock_pkey; +} +# Check the table from the perspective of s1. +# +# Besides the contents, we also check that relfilenode has changed. + +# Have each session write the contents into a table and use FULL JOIN to check +# if the outputs are identical. +step check1 +{ + INSERT INTO relfilenodes(node) + SELECT relfilenode FROM pg_class WHERE relname='repack_deadlock'; + + SELECT count(DISTINCT node) FROM relfilenodes; + + SELECT i, j FROM repack_deadlock ORDER BY i, j; + + INSERT INTO data_s1(i, j) + SELECT i, j FROM repack_deadlock; + + SELECT count(*) + FROM data_s1 d1 FULL JOIN data_s2 d2 USING (i, j) + WHERE d1.i ISNULL OR d2.i ISNULL; +} +teardown +{ + SELECT injection_points_detach('repack-concurrently-before-lock'); +} + +session s2 +# Change the existing data. UPDATE changes both key and non-key columns. Also +# update one row twice to test whether tuple version generated by this session +# can be found. +step add_column +{ + alter table repack_deadlock add column noise text; +} + +step wakeup_before_lock +{ + SELECT injection_points_wakeup('repack-concurrently-before-lock'); +} + +# Test if data changes introduced while one session is performing REPACK +# CONCURRENTLY find their way into the table. +permutation + wait_before_lock + add_column + wakeup_before_lock + check1 -- 2.47.3 --kdrcpfmkbkc4lqhu Content-Type: text/x-diff; charset=utf-8 Content-Disposition: attachment; filename="0002-Publish-list-of-tables-being-repacked-in-shared-memo.patch" Content-Transfer-Encoding: 8bit ^ permalink raw reply [nested|flat] 10+ messages in thread
end of thread, other threads:[~2026-04-01 15:35 UTC | newest] Thread overview: 10+ messages (download: mbox mbox.gz follow: Atom feed) -- links below jump to the message on this page -- 2019-12-13 02:54 [PATCH v17 1/2] vacuum errcontext to show block being processed Justin Pryzby <pryzbyj@telsasoft.com> 2019-12-13 02:54 [PATCH v22 1/3] vacuum errcontext to show block being processed Justin Pryzby <pryzbyj@telsasoft.com> 2019-12-13 02:54 [PATCH v9 2/3] vacuum errcontext to show block being processed Justin Pryzby <pryzbyj@telsasoft.com> 2019-12-13 02:54 [PATCH v27 1/5] vacuum errcontext to show block being processed Justin Pryzby <pryzbyj@telsasoft.com> 2019-12-13 02:54 [PATCH v28 1/5] vacuum errcontext to show block being processed Justin Pryzby <pryzbyj@telsasoft.com> 2019-12-13 02:54 [PATCH v23 1/3] vacuum errcontext to show block being processed Justin Pryzby <pryzbyj@telsasoft.com> 2019-12-13 02:54 [PATCH v8 4/5] vacuum errcontext to show block being processed Justin Pryzby <pryzbyj@telsasoft.com> 2019-12-13 02:54 [PATCH v25 1/4] vacuum errcontext to show block being processed Justin Pryzby <pryzbyj@telsasoft.com> 2019-12-13 02:54 [PATCH v24 1/4] vacuum errcontext to show block being processed Justin Pryzby <pryzbyj@telsasoft.com> 2026-04-01 15:35 [PATCH 1/2] Error out any process that would block at REPACK Antonin Houska <ah@cybertec.at>
This inbox is served by agora; see mirroring instructions for how to clone and mirror all data and code used for this inbox