public inbox for [email protected]
help / color / mirror / Atom feedFrom: Zhijie Hou (Fujitsu) <[email protected]>
To: Zhijie Hou (Fujitsu) <[email protected]>
Cc: shveta malik <[email protected]>
Cc: Hayato Kuroda (Fujitsu) <[email protected]>
Cc: PostgreSQL Hackers <[email protected]>
Cc: Tomas Vondra <[email protected]>
Cc: Dilip Kumar <[email protected]>
Cc: Andrei Lepikhov <[email protected]>
Cc: wenhui qiu <[email protected]>
Cc: Amit Kapila <[email protected]>
Cc: Peter Smith <[email protected]>
Subject: RE: Parallel Apply
Date: Thu, 9 Jul 2026 08:08:31 +0000
Message-ID: <TY4PR01MB177180394D8E9C6D263E117E494FE2@TY4PR01MB17718.jpnprd01.prod.outlook.com> (raw)
In-Reply-To: <TY4PR01MB1771875A7A64B139EF8C8D05C94E62@TY4PR01MB17718.jpnprd01.prod.outlook.com>
References: <CAJpy0uDCxK2wiM-uMYSMKkrp+pYOHxE5ns4ihFvht=JmytVJ3A@mail.gmail.com>
<CAJpy0uByjHr2mEDPPghzvtRkQt__BnqALs5fgkSwa5DALT=Q6w@mail.gmail.com>
<TY4PR01MB17718A4AB7F564C8DB0D6F7DC94102@TY4PR01MB17718.jpnprd01.prod.outlook.com>
<CAHut+Pt=C9Z5PSkELKaTOh3AfjoUVf__FU2sQCFs8L60Qvfxhw@mail.gmail.com>
<TY4PR01MB1771875A7A64B139EF8C8D05C94E62@TY4PR01MB17718.jpnprd01.prod.outlook.com>
On Monday, June 15, 2026 7:20 PM Zhijie Hou (Fujitsu) <[email protected]> wrote:
>
> Attaching the new version patch set.
>
> Apart from addressing above comments, I also did the following changes:
>
> 0003:
>
> - Fix detection of dependencies on toasted columns by properly copying
> column
> status.
> - Ensure leader correctly handles table-wide dependencies when applying a
> transaction directly.
> - Prevent infinite loop by skipping self-dependencies when building the
> dependency list.
> - Expand test coverage for dependency tracking.
> - Improve code readability through refactoring.
> - Enhance comments for better documentation.
>
> 0008:
> 0009:
>
> - Complete most TODOs and finalize the design to cover more cases. Add
> extensive
> comments to explain the design.
>
> - Add more tests to cover the new code paths.
I've further reviewed and improved the patch recently; here is the updated V21 patch
set with the following changes:
0003: Fix missing dependency check for replica identity FULL.
0004: Add more test coverage for replica identity FULL.
0007: Improve parallel apply safety check to handle concurrent table schema
changes.
0009: Fix missing dependency check when both sides of a foreign key are
partitioned tables, and add more test coverage for partitioned tables.
Best Regards,
Hou zj
Attachments:
[application/octet-stream] v21-0001-Introduce-a-shared-hash-table-to-store-paralleli.patch (9.1K, ../TY4PR01MB177180394D8E9C6D263E117E494FE2@TY4PR01MB17718.jpnprd01.prod.outlook.com/2-v21-0001-Introduce-a-shared-hash-table-to-store-paralleli.patch)
download | inline diff:
From 563e590c157c9db561a9beaf2a4a4a93c9e4576a Mon Sep 17 00:00:00 2001
From: Zhijie Hou <[email protected]>
Date: Wed, 27 May 2026 11:31:35 +0800
Subject: [PATCH v21 1/9] Introduce a shared hash table to store parallelized
transactions
This commit introduces a shared hash table for tracking parallelized
transactions. The hash table uses transaction IDs as keys; no values are stored.
The hash entries are inserted with a remote XID when the leader bypasses remote
transactions to parallel apply workers. Entries are deleted when parallel
workers are committed to corresponding transactions.
Any worker can use this hash table to check whether a transaction is still
in progress:
- If an entry exists, the transaction is still in progress.
- If no entry exists, the transaction has committed.
Later patches will use this hash table to support dependency waiting and
commit order preservation. When transaction A depends on transaction B
(or must commit after B), the worker will wait for transaction B's entry
to be removed before committing transaction A.
Author: Zhijie Hou <[email protected]>
Author: Hayato Kuroda <[email protected]>
---
.../replication/logical/applyparallelworker.c | 90 ++++++++++++++++++-
.../utils/activity/wait_event_names.txt | 1 +
src/include/replication/worker_internal.h | 9 ++
src/include/storage/lwlocklist.h | 1 +
src/tools/pgindent/typedefs.list | 1 +
5 files changed, 101 insertions(+), 1 deletion(-)
diff --git a/src/backend/replication/logical/applyparallelworker.c b/src/backend/replication/logical/applyparallelworker.c
index 012d55e9d3d..5eb0679e850 100644
--- a/src/backend/replication/logical/applyparallelworker.c
+++ b/src/backend/replication/logical/applyparallelworker.c
@@ -223,10 +223,36 @@ typedef struct ParallelApplyWorkerEntry
/*
* A hash table used to cache the state of streaming transactions being applied
- * by the parallel apply workers.
+ * by the parallel apply workers. Entries are of type ParallelApplyWorkerEntry.
*/
static HTAB *ParallelApplyTxnHash = NULL;
+/* An entry in the parallelized_txns shared hash table */
+typedef struct ParallelizedTxnEntry
+{
+ TransactionId xid; /* Hash key, remote transaction ID */
+} ParallelizedTxnEntry;
+
+/*
+ * A hash table used to track the parallelized remote transactions that could be
+ * depended on by other transactions. Entries are of type ParallelizedTxnEntry.
+ *
+ * dshash is used to enable dynamic shared memory allocation based on the number
+ * of transactions being applied in parallel.
+ */
+static dsa_area *parallel_apply_dsa_area = NULL;
+static dshash_table *parallelized_txns = NULL;
+
+/* Parameters for the parallelized_txns shared hash table */
+static const dshash_parameters dsh_params = {
+ sizeof(TransactionId),
+ sizeof(ParallelizedTxnEntry),
+ dshash_memcmp,
+ dshash_memhash,
+ dshash_memcpy,
+ LWTRANCHE_PARALLEL_APPLY_DSA
+};
+
/*
* A list (pool) of active parallel apply workers. The information for
* the new worker is added to the list after successfully launching it. The
@@ -260,6 +286,8 @@ static List *subxactlist = NIL;
static void pa_free_worker_info(ParallelApplyWorkerInfo *winfo);
static ParallelTransState pa_get_xact_state(ParallelApplyWorkerShared *wshared);
static PartialFileSetState pa_get_fileset_state(void);
+static void pa_attach_parallelized_txn_hash(dsa_handle *pa_dsa_handle,
+ dshash_table_handle *pa_dshash_handle);
/*
* Returns true if it is OK to start a parallel apply worker, false otherwise.
@@ -337,6 +365,15 @@ pa_setup_dsm(ParallelApplyWorkerInfo *winfo)
shm_mq *mq;
Size queue_size = DSM_QUEUE_SIZE;
Size error_queue_size = DSM_ERROR_QUEUE_SIZE;
+ dsa_handle parallel_apply_dsa_handle;
+ dshash_table_handle parallelized_txns_handle;
+
+ pa_attach_parallelized_txn_hash(¶llel_apply_dsa_handle,
+ ¶llelized_txns_handle);
+
+ if (parallel_apply_dsa_handle == DSA_HANDLE_INVALID ||
+ parallelized_txns_handle == DSHASH_HANDLE_INVALID)
+ return false;
/*
* Estimate how much shared memory we need.
@@ -372,6 +409,8 @@ pa_setup_dsm(ParallelApplyWorkerInfo *winfo)
pg_atomic_init_u32(&(shared->pending_stream_count), 0);
shared->last_commit_end = InvalidXLogRecPtr;
shared->fileset_state = FS_EMPTY;
+ shared->parallel_apply_dsa_handle = parallel_apply_dsa_handle;
+ shared->parallelized_txns_handle = parallelized_txns_handle;
shm_toc_insert(toc, PARALLEL_APPLY_KEY_SHARED, shared);
@@ -876,6 +915,8 @@ ParallelApplyWorkerMain(Datum main_arg)
shm_mq *mq;
shm_mq_handle *mqh;
shm_mq_handle *error_mqh;
+ dsa_handle pa_dsa_handle;
+ dshash_table_handle pa_dshash_handle;
ReplOriginId originid;
int worker_slot = DatumGetInt32(main_arg);
char originname[NAMEDATALEN];
@@ -962,6 +1003,8 @@ ParallelApplyWorkerMain(Datum main_arg)
InitializingApplyWorker = false;
+ pa_attach_parallelized_txn_hash(&pa_dsa_handle, &pa_dshash_handle);
+
/* Setup replication origin tracking. */
StartTransactionCommand();
ReplicationOriginNameForLogicalRep(MySubscription->oid, InvalidOid,
@@ -1656,3 +1699,48 @@ pa_xact_finish(ParallelApplyWorkerInfo *winfo, XLogRecPtr remote_lsn)
pa_free_worker(winfo);
}
+
+/*
+ * Attach to the shared hash table for parallelized transactions.
+ */
+static void
+pa_attach_parallelized_txn_hash(dsa_handle *pa_dsa_handle,
+ dshash_table_handle *pa_dshash_handle)
+{
+ MemoryContext oldctx;
+
+ if (parallelized_txns)
+ {
+ Assert(parallel_apply_dsa_area);
+ *pa_dsa_handle = dsa_get_handle(parallel_apply_dsa_area);
+ *pa_dshash_handle = dshash_get_hash_table_handle(parallelized_txns);
+ return;
+ }
+
+ /* Be sure any local memory allocated by DSA routines is persistent. */
+ oldctx = MemoryContextSwitchTo(ApplyContext);
+
+ if (am_leader_apply_worker())
+ {
+ /* Initialize dynamic shared hash table for parallelized transactions */
+ parallel_apply_dsa_area = dsa_create(LWTRANCHE_PARALLEL_APPLY_DSA);
+ dsa_pin(parallel_apply_dsa_area);
+ dsa_pin_mapping(parallel_apply_dsa_area);
+ parallelized_txns = dshash_create(parallel_apply_dsa_area, &dsh_params, NULL);
+
+ /* Store handles in shared memory for other backends to use. */
+ *pa_dsa_handle = dsa_get_handle(parallel_apply_dsa_area);
+ *pa_dshash_handle = dshash_get_hash_table_handle(parallelized_txns);
+ }
+ else if (am_parallel_apply_worker())
+ {
+ /* Attach to existing dynamic shared hash table. */
+ parallel_apply_dsa_area = dsa_attach(MyParallelShared->parallel_apply_dsa_handle);
+ dsa_pin_mapping(parallel_apply_dsa_area);
+ parallelized_txns = dshash_attach(parallel_apply_dsa_area, &dsh_params,
+ MyParallelShared->parallelized_txns_handle,
+ NULL);
+ }
+
+ MemoryContextSwitchTo(oldctx);
+}
diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt
index 560659f9568..1339c33262f 100644
--- a/src/backend/utils/activity/wait_event_names.txt
+++ b/src/backend/utils/activity/wait_event_names.txt
@@ -417,6 +417,7 @@ XactSLRU "Waiting to access the transaction status SLRU cache."
ParallelVacuumDSA "Waiting for parallel vacuum dynamic shared memory allocation."
AioUringCompletion "Waiting for another process to complete IO via io_uring."
ShmemIndex "Waiting to find or allocate space in shared memory."
+ParallelApplyDSA "Waiting for parallel apply dynamic shared memory allocation."
# No "ABI_compatibility" region here as WaitEventLWLock has its own C code.
diff --git a/src/include/replication/worker_internal.h b/src/include/replication/worker_internal.h
index 745b7d9e969..e60e0c3d6d7 100644
--- a/src/include/replication/worker_internal.h
+++ b/src/include/replication/worker_internal.h
@@ -15,6 +15,7 @@
#include "access/xlogdefs.h"
#include "catalog/pg_subscription.h"
#include "datatype/timestamp.h"
+#include "lib/dshash.h"
#include "miscadmin.h"
#include "replication/logicalrelation.h"
#include "replication/walreceiver.h"
@@ -196,6 +197,14 @@ typedef struct ParallelApplyWorkerShared
*/
PartialFileSetState fileset_state;
FileSet fileset;
+
+ /*
+ * DSA handle for parallel apply workers, along with the handle for the
+ * shared hash table allocated under it. The hash table stores parallelized
+ * transaction information.
+ */
+ dsa_handle parallel_apply_dsa_handle;
+ dshash_table_handle parallelized_txns_handle;
} ParallelApplyWorkerShared;
/*
diff --git a/src/include/storage/lwlocklist.h b/src/include/storage/lwlocklist.h
index d7eb648bd27..c5f2921b703 100644
--- a/src/include/storage/lwlocklist.h
+++ b/src/include/storage/lwlocklist.h
@@ -140,3 +140,4 @@ PG_LWLOCKTRANCHE(XACT_SLRU, XactSLRU)
PG_LWLOCKTRANCHE(PARALLEL_VACUUM_DSA, ParallelVacuumDSA)
PG_LWLOCKTRANCHE(AIO_URING_COMPLETION, AioUringCompletion)
PG_LWLOCKTRANCHE(SHMEM_INDEX, ShmemIndex)
+PG_LWLOCKTRANCHE(PARALLEL_APPLY_DSA, ParallelApplyDSA)
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index ffb413ab612..32524990758 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -2164,6 +2164,7 @@ ParallelHashJoinBatch
ParallelHashJoinBatchAccessor
ParallelHashJoinState
ParallelIndexScanDesc
+ParallelizedTxnEntry
ParallelSlot
ParallelSlotArray
ParallelSlotResultHandler
--
2.43.0
[application/octet-stream] v21-0009-Support-dependency-tracking-via-local-foreign-ke.patch (49.7K, ../TY4PR01MB177180394D8E9C6D263E117E494FE2@TY4PR01MB17718.jpnprd01.prod.outlook.com/3-v21-0009-Support-dependency-tracking-via-local-foreign-ke.patch)
download | inline diff:
From b959f17179689938195efde244fbecff4c9d32ba Mon Sep 17 00:00:00 2001
From: Zhijie Hou <[email protected]>
Date: Tue, 7 Jul 2026 19:18:05 +0800
Subject: [PATCH v21 9/9] Support dependency tracking via local foreign keys
This patch tracks foreign key dependencies to prevent constraint violations.
Consider a referencing table FK_TABLE (a INT REFERENCES PK_TABLE) and a
referenced table PK_TABLE (a INT PRIMARY KEY):
TX-1: INSERT row (1) INTO PK_TABLE
TX-2: INSERT row (1) INTO FK_TABLE
TX-2 must wait for TX-1; otherwise, a foreign key violation occurs.
Similarly, for deletions:
TX-1: DELETE row (1) FROM FK_TABLE
TX-2: DELETE row (1) FROM PK_TABLE
TX-2 must wait for TX-1 to avoid violating the foreign key constraint.
Therefore, we record new tuples on the referenced table (PK_TABLE) and check
for dependencies when processing new tuples on the referencing table
(FK_TABLE). Conversely, we record old tuples deleted from the referencing
table and require deletions on the referenced table to depend on them.
Similar to unique key dependencies, for delete-delete cases we only track
foreign key dependencies on columns that are part of the replica identity
keys on the publisher. This may lead to false positives but is acceptable
given the trade-off.
---
.../replication/logical/applyparallelworker.c | 26 ++
src/backend/replication/logical/relation.c | 384 +++++++++++++++++
src/backend/replication/logical/worker.c | 395 +++++++++++++++++-
src/include/replication/logicalrelation.h | 30 ++
src/test/subscription/t/050_parallel_apply.pl | 330 +++++++++++++++
5 files changed, 1150 insertions(+), 15 deletions(-)
diff --git a/src/backend/replication/logical/applyparallelworker.c b/src/backend/replication/logical/applyparallelworker.c
index 1f1a75b5019..733879d1edc 100644
--- a/src/backend/replication/logical/applyparallelworker.c
+++ b/src/backend/replication/logical/applyparallelworker.c
@@ -218,6 +218,32 @@
* replica identity keys. This could be useful for users who prefer higher
* parallelism and experience few conflicts.
*
+ * Additionally, foreign key dependencies are also tracked to prevent constraint
+ * violations. Consider a referencing table FK_TABLE (a INT REFERENCES PK_TABLE)
+ * and a referenced table PK_TABLE (a INT PRIMARY KEY):
+ *
+ * TX-1: INSERT row (1) INTO PK_TABLE
+ * TX-2: INSERT row (1) INTO FK_TABLE
+ *
+ * TX-2 must wait for TX-1; otherwise, a foreign key violation occurs.
+ *
+ * Similarly, for deletions:
+ *
+ * TX-1: DELETE row (1) FROM FK_TABLE
+ * TX-2: DELETE row (1) FROM PK_TABLE
+ *
+ * TX-2 must wait for TX-1 to avoid violating the foreign key constraint.
+ *
+ * Therefore, we record new tuples on the referenced table (PK_TABLE) and check
+ * for dependencies when processing new tuples on the referencing table
+ * (FK_TABLE). Conversely, we record old tuples deleted from the referencing
+ * table and require deletions on the referenced table to depend on them.
+ *
+ * Similar to unique key dependencies, for delete-delete cases we only track
+ * foreign key dependencies on columns that are part of the replica identity
+ * keys on the publisher. This may lead to false positives but is acceptable
+ * given the trade-off.
+ *
* Note that the local unique key could change after dependency checking and
* before applying the change. However, to centralize tracking and keep it
* simple, we still perform this check only in the leader apply worker. This is
diff --git a/src/backend/replication/logical/relation.c b/src/backend/replication/logical/relation.c
index 1572990ee18..cdafa10beb3 100644
--- a/src/backend/replication/logical/relation.c
+++ b/src/backend/replication/logical/relation.c
@@ -20,7 +20,11 @@
#include "access/amapi.h"
#include "access/genam.h"
#include "access/table.h"
+#include "catalog/heap.h"
#include "catalog/namespace.h"
+#include "catalog/partition.h"
+#include "catalog/pg_constraint.h"
+#include "catalog/pg_inherits.h"
#include "catalog/pg_proc.h"
#include "catalog/pg_subscription_rel.h"
#include "catalog/pg_trigger.h"
@@ -31,6 +35,7 @@
#include "replication/logicalrelation.h"
#include "replication/worker_internal.h"
#include "rewrite/rewriteHandler.h"
+#include "utils/fmgroids.h"
#include "utils/inval.h"
#include "utils/lsyscache.h"
#include "utils/syscache.h"
@@ -62,6 +67,10 @@ typedef struct LogicalRepPartMapEntry
static Oid FindLogicalRepLocalIndex(Relation localrel, LogicalRepRelation *remoterel,
AttrMap *attrMap);
+static void logicalrep_map_attnums_by_name(Oid src_reloid, Oid dst_reloid,
+ int nkeys,
+ const AttrNumber *src_attnums,
+ AttrNumber *dst_attnums);
/*
* Relcache invalidation callback for our relation map cache.
@@ -89,6 +98,7 @@ logicalrep_relmap_invalidate_cb(Datum arg, Oid reloid)
entry->localrelvalid = false;
entry->parallel_safety_valid = false;
entry->local_unique_indexes_valid = false;
+ entry->local_fkeys_valid = false;
hash_seq_term(&status);
break;
}
@@ -106,6 +116,7 @@ logicalrep_relmap_invalidate_cb(Datum arg, Oid reloid)
entry->localrelvalid = false;
entry->parallel_safety_valid = false;
entry->local_unique_indexes_valid = false;
+ entry->local_fkeys_valid = false;
}
}
}
@@ -152,6 +163,33 @@ free_local_unique_indexes(LogicalRepRelMapEntry *entry)
entry->local_unique_indexes = NIL;
}
+/*
+ * Release local foreign key lists.
+ */
+static void
+free_local_fkeys(LogicalRepRelMapEntry *entry)
+{
+ Assert(am_leader_apply_worker());
+
+ foreach_ptr(LogicalRepSubFKey, fkinfo, entry->local_fkeys)
+ {
+ list_free(fkinfo->fkattnums_new);
+ list_free(fkinfo->fkattnums_old);
+ }
+
+ foreach_ptr(LogicalRepSubRefKey, refinfo, entry->local_refkeys)
+ {
+ list_free(refinfo->refattnums_new);
+ list_free(refinfo->refattnums_old);
+ }
+
+ list_free_deep(entry->local_fkeys);
+ list_free_deep(entry->local_refkeys);
+
+ entry->local_fkeys = NIL;
+ entry->local_refkeys = NIL;
+}
+
/*
* Free the entry of a relation map cache.
*/
@@ -182,6 +220,9 @@ logicalrep_relmap_free_entry(LogicalRepRelMapEntry *entry)
if (entry->local_unique_indexes != NIL)
free_local_unique_indexes(entry);
+
+ if (entry->local_fkeys != NIL || entry->local_refkeys != NIL)
+ free_local_fkeys(entry);
}
/*
@@ -432,6 +473,7 @@ logicalrep_rel_open(LogicalRepRelId remoteid, LOCKMODE lockmode)
entry->localrelvalid = false;
entry->parallel_safety_valid = false;
entry->local_unique_indexes_valid = false;
+ entry->local_fkeys_valid = false;
}
else if (!entry->localrelvalid)
{
@@ -793,6 +835,348 @@ logicalrep_build_dependent_unique_indexes(LogicalRepRelMapEntry *entry)
entry->local_unique_indexes_valid = true;
}
+/*
+ * Return all local relation map entries that match the given list of local
+ * relation OIDs.
+ */
+static List *
+logicalrep_get_local_relentries(List *localrelids)
+{
+ HASH_SEQ_STATUS status;
+ LogicalRepRelMapEntry *entry;
+ List *entries = NIL;
+
+ if (LogicalRepRelMap == NULL)
+ return NIL;
+
+ hash_seq_init(&status, LogicalRepRelMap);
+ while ((entry = (LogicalRepRelMapEntry *) hash_seq_search(&status)) != NULL)
+ {
+ if (!OidIsValid(entry->localreloid))
+ continue;
+
+ if (list_member_oid(localrelids, entry->localreloid))
+ entries = lappend(entries, entry);
+ }
+
+ return entries;
+}
+
+/*
+ * Map key columns from src_reloid to dst_reloid by attribute name.
+ */
+static void
+logicalrep_map_attnums_by_name(Oid src_reloid, Oid dst_reloid,
+ int nkeys,
+ const AttrNumber *src_attnums,
+ AttrNumber *dst_attnums)
+{
+ if (src_reloid == dst_reloid)
+ {
+ memcpy(dst_attnums, src_attnums, nkeys * sizeof(AttrNumber));
+ return;
+ }
+
+ for (int i = 0; i < nkeys; i++)
+ {
+ char *attname;
+ AttrNumber attnum;
+
+ attname = get_attname(src_reloid, src_attnums[i], false);
+ attnum = get_attnum(dst_reloid, attname);
+ pfree(attname);
+
+ Assert(AttributeNumberIsValid(attnum));
+
+ dst_attnums[i] = attnum;
+ }
+}
+
+/*
+ * Check whether the table has any foreign key triggers of the given type
+ * (trig_type) enabled in replica mode.
+ */
+static bool
+fkey_trigger_enabled_in_replica(Relation rel, Oid conoid, int trig_type)
+{
+ TriggerDesc *trigdesc = rel->trigdesc;
+
+ if (trigdesc == NULL)
+ return false;
+
+ for (int i = 0; i < trigdesc->numtriggers; i++)
+ {
+ Trigger *trig = &trigdesc->triggers[i];
+
+ /* Keep only PK-side action triggers for this FK constraint */
+ if (trig->tgconstraint != conoid ||
+ RI_FKey_trigger_type(trig->tgfoid) != trig_type)
+ continue;
+
+ /* In replica mode, only REPLICA/ALWAYS triggers can fire. */
+ if (trig->tgenabled == TRIGGER_FIRES_ON_REPLICA ||
+ trig->tgenabled == TRIGGER_FIRES_ALWAYS)
+ return true;
+ }
+
+ return false;
+}
+
+/*
+ * Build lists of foreign key and referenced primary key remote columns. Both
+ * lists are ordered according to the referenced table's remote column order.
+ */
+static void
+build_fk_remote_attnums(LogicalRepRelMapEntry *fkentry,
+ LogicalRepRelMapEntry *refentry,
+ int nkeys,
+ const AttrNumber *fk_conkey,
+ const AttrNumber *ref_confkey,
+ int trig_type,
+ List **fkattnums_new,
+ List **fkattnums_old)
+{
+ int pkatt = -1;
+
+ Assert(trig_type == RI_TRIGGER_FK || trig_type == RI_TRIGGER_PK);
+
+ /*
+ * Traverse the referenced table's remote replica identity columns in order,
+ * and for each, find the corresponding foreign key column that matches it.
+ */
+ while ((pkatt = bms_next_member(refentry->remoterel.attkeys, pkatt)) >= 0)
+ {
+ for (int i = 0; i < nkeys; i++)
+ {
+ AttrNumber fk_local_attnum = fk_conkey[i];
+ AttrNumber ref_local_attnum = ref_confkey[i];
+ int fk_remote_attnum;
+ int ref_remote_attnum;
+ int mapped_attnum;
+
+ fk_remote_attnum = fkentry->attrmap->attnums[AttrNumberGetAttrOffset(fk_local_attnum)];
+ ref_remote_attnum = refentry->attrmap->attnums[AttrNumberGetAttrOffset(ref_local_attnum)];
+
+ /* Skip if not the current traversed column */
+ if (ref_remote_attnum != pkatt)
+ continue;
+
+ /* Skip columns that are unavailable in the remote table */
+ if (ref_remote_attnum < 0 || fk_remote_attnum < 0)
+ continue;
+
+ if (trig_type == RI_TRIGGER_FK)
+ mapped_attnum = fk_remote_attnum + 1;
+ else if (trig_type == RI_TRIGGER_PK)
+ mapped_attnum = ref_remote_attnum + 1;
+
+ *fkattnums_new = lappend_int(*fkattnums_new, mapped_attnum);
+
+ /* Old tuple contains only replica identity columns. */
+ if (bms_is_member(fk_remote_attnum, fkentry->remoterel.attkeys) &&
+ bms_is_member(ref_remote_attnum, refentry->remoterel.attkeys))
+ *fkattnums_old = lappend_int(*fkattnums_old, mapped_attnum);
+
+ break;
+ }
+ }
+}
+
+/*
+ * Collect foreign keys for dependency tracking.
+ *
+ * This function collects both foreign keys in the referencing table and the
+ * primary key in the referenced table, as both are needed for dependency
+ * tracking (see applyparallelworker.c for details).
+ *
+ * For referencing-side foreign keys, we cannot use the referencing column
+ * numbers directly, because their order may differ from the primary key column
+ * order in the referenced table. To ensure consistency, we store a list of
+ * column numbers in the order of the referenced table's remote columns. The
+ * dependency tracking function will traverse this list to build the hash key.
+ * For referenced-side primary keys, we preserve the original column order.
+ *
+ * For one foreign key constraint on the table, We collect two set of column
+ * numbers for both the referencing and referenced keys:
+ *
+ * - One for INSERT-INSERT dependency tracking, using the new tuple. This
+ * includes all remote columns referenced by the foreign key, as the new tuple
+ * contains all remote values.
+ *
+ * - One for DELETE-DELETE dependency tracking, using the old tuple. This
+ * includes only columns that are part of the replica identity key, because
+ * the old tuple of an UPDATE or DELETE contains only replica identity key
+ * columns. Any other columns would be missing and unavailable for dependency
+ * tracking.
+ *
+ * If there are multiple foreign keys referencing other tables, or if the
+ * primary key is referenced by multiple foreign keys, we will have multiple
+ * column number sets. One might think that for a referenced primary key, we
+ * only need to store a single column bitmap (since only one primary key
+ * exists). However, the referencing side may not have all those column values
+ * available (columns outside the replica identity). Therefore, we store
+ * multiple primary key column sets, each including only the columns that are
+ * available on both sides. This ensures that the remote values on both the
+ * referenced and referencing tables can compute the same hash value.
+ */
+void
+logicalrep_build_dependent_fkeys(LogicalRepRelMapEntry *entry)
+{
+ int trig_type;
+ Relation fkeyRel;
+ SysScanDesc fkeyScan;
+ HeapTuple tuple;
+ Oid relid = RelationGetRelid(entry->localrel);
+
+ Assert(OidIsValid(relid));
+
+ if (entry->local_fkeys_valid)
+ return;
+
+ free_local_fkeys(entry);
+
+ fkeyRel = table_open(ConstraintRelationId, AccessShareLock);
+
+ fkeyScan = systable_beginscan(fkeyRel, InvalidOid, false,
+ NULL, 0, NULL);
+
+ while (HeapTupleIsValid(tuple = systable_getnext(fkeyScan)))
+ {
+ Form_pg_constraint con = (Form_pg_constraint) GETSTRUCT(tuple);
+ List *fkentries;
+ AttrNumber conkey[INDEX_MAX_KEYS] = {0};
+ AttrNumber confkey[INDEX_MAX_KEYS] = {0};
+ AttrNumber *key_to_map;
+ int numfks;
+ MemoryContext oldctx;
+ List *all_parts;
+ Oid otherrelid;
+
+ /* Not a foreign key */
+ if (con->contype != CONSTRAINT_FOREIGN)
+ continue;
+
+ /* Skip if FK enforcement is disabled */
+ if (!con->conenforced)
+ continue;
+
+ /*
+ * Deferrable foreign keys do not need tracking, as they won't cause
+ * constraint violations as long as commit order is preserved.
+ */
+ if (con->condeferrable && con->condeferred)
+ continue;
+
+ if (con->confrelid == relid)
+ {
+ trig_type = RI_TRIGGER_PK;
+ otherrelid = con->conrelid;
+ key_to_map = conkey;
+ }
+ else if (con->conrelid == relid)
+ {
+ trig_type = RI_TRIGGER_FK;
+ otherrelid = con->confrelid;
+ key_to_map = confkey;
+ }
+ else
+ continue;
+
+ /*
+ * Skip when no replica-mode ON DELETE/ON UPDATE violation trigger can
+ * fire for this FK on the referenced table.
+ */
+ if (!fkey_trigger_enabled_in_replica(entry->localrel, con->oid,
+ trig_type))
+ continue;
+
+ DeconstructFkConstraintRow(tuple, &numfks, conkey, confkey,
+ NULL, NULL, NULL, NULL, NULL);
+
+ /*
+ * Collect all partitions of the referenced (or referencing) table if
+ * it's partitioned, and cache foreign key data between the current
+ * table and all of its partitions.
+ *
+ * This is necessary when both the referencing and referenced sides are
+ * partitioned tables, as pg_constraint does not record foreign key
+ * relationships between child partitions. We must search them
+ * explicitly.
+ *
+ * No lock is needed on each partition because we already hold a lock on
+ * the referencing table, preventing concurrent alterations that would
+ * affect foreign key dependencies.
+ */
+ all_parts = find_all_inheritors(otherrelid, NoLock, NULL);
+ fkentries = logicalrep_get_local_relentries(all_parts);
+
+ foreach_ptr(LogicalRepRelMapEntry, fkentry, fkentries)
+ {
+ AttrNumber mapped_conkey[INDEX_MAX_KEYS] = {0};
+
+ /*
+ * Skip if the referencing table is not published or has not
+ * replicated any changes.
+ */
+ if (!fkentry->attrmap)
+ continue;
+
+ /*
+ * Map column numbers from the parent to the corresponding child
+ * columns.
+ */
+ logicalrep_map_attnums_by_name(otherrelid,
+ fkentry->localreloid,
+ numfks,
+ key_to_map,
+ mapped_conkey);
+
+ oldctx = MemoryContextSwitchTo(LogicalRepRelMapContext);
+
+ if (trig_type == RI_TRIGGER_PK)
+ {
+ LogicalRepSubRefKey *refinfo;
+
+ refinfo = palloc0_object(LogicalRepSubRefKey);
+ refinfo->conoid = con->oid;
+ refinfo->fk_remoteid = fkentry->remoterel.remoteid;
+
+ build_fk_remote_attnums(fkentry, entry, numfks, mapped_conkey,
+ confkey, trig_type,
+ &refinfo->refattnums_new,
+ &refinfo->refattnums_old);
+
+ entry->local_refkeys = lappend(entry->local_refkeys, refinfo);
+ }
+ else if (trig_type == RI_TRIGGER_FK)
+ {
+ LogicalRepSubFKey *fkinfo;
+
+ fkinfo = palloc0_object(LogicalRepSubFKey);
+ fkinfo->conoid = con->oid;
+ fkinfo->ref_remoteid = fkentry->remoterel.remoteid;
+
+ build_fk_remote_attnums(entry, fkentry, numfks, conkey,
+ mapped_conkey, trig_type,
+ &fkinfo->fkattnums_new,
+ &fkinfo->fkattnums_old);
+
+ entry->local_fkeys = lappend(entry->local_fkeys, fkinfo);
+ }
+
+ MemoryContextSwitchTo(oldctx);
+ }
+
+ list_free(fkentries);
+ }
+
+ systable_endscan(fkeyScan);
+
+ table_close(fkeyRel, AccessShareLock);
+
+ entry->local_fkeys_valid = true;
+}
+
/*
* Partition cache: look up partition LogicalRepRelMapEntry's
*
diff --git a/src/backend/replication/logical/worker.c b/src/backend/replication/logical/worker.c
index 3864c95f930..08b4ddb094a 100644
--- a/src/backend/replication/logical/worker.c
+++ b/src/backend/replication/logical/worker.c
@@ -255,6 +255,7 @@
#include "access/twophase.h"
#include "access/xact.h"
#include "catalog/indexing.h"
+#include "catalog/namespace.h"
#include "catalog/pg_inherits.h"
#include "catalog/pg_subscription.h"
#include "catalog/pg_subscription_rel.h"
@@ -265,6 +266,7 @@
#include "executor/execPartition.h"
#include "libpq/pqformat.h"
#include "miscadmin.h"
+#include "nodes/makefuncs.h"
#include "optimizer/optimizer.h"
#include "parser/parse_relation.h"
#include "pgstat.h"
@@ -584,7 +586,10 @@ static ApplySubXactData subxact_data = {0, 0, InvalidTransactionId, NULL};
typedef enum LogicalRepKeyKind
{
LOGICALREP_KEY_REPLICA_IDENTITY,
- LOGICALREP_KEY_LOCAL_UNIQUE
+ LOGICALREP_KEY_LOCAL_UNIQUE,
+ LOGICALREP_KEY_FOREIGN_KEY,
+ LOGICALREP_KEY_REFERENCED_KEY,
+ LOGICALREP_ALL_KEYS,
} LogicalRepKeyKind;
/* Hash table key for replica_identity_table */
@@ -950,6 +955,26 @@ append_xid_dependency(TransactionId xid, List **depends_on_xids)
*depends_on_xids = lappend_xid(*depends_on_xids, xid);
}
+static char *
+get_dependency_type_str(LogicalRepKeyKind kind)
+{
+ switch (kind)
+ {
+ case LOGICALREP_KEY_REPLICA_IDENTITY:
+ return "replica identity";
+ case LOGICALREP_KEY_LOCAL_UNIQUE:
+ return "local unique key";
+ case LOGICALREP_KEY_FOREIGN_KEY:
+ return "foreign key";
+ case LOGICALREP_KEY_REFERENCED_KEY:
+ return "referenced key";
+ case LOGICALREP_ALL_KEYS:
+ return "all keys";
+ }
+
+ return "???";
+}
+
/*
* Common function for checking dependency by using the key. Used by both
* check_and_record_ri_dependency and check_and_record_local_key_dependency.
@@ -982,9 +1007,8 @@ check_and_record_key_dependency(ReplicaIdentityKey *key,
if (rientry && has_active_key_dependency(rientry, true))
{
elog(DEBUG1,
- key->kind == LOGICALREP_KEY_REPLICA_IDENTITY ?
- "found conflicting replica identity change on table %u from %u" :
- "found conflicting local unique key change on table %u from %u",
+ "found conflicting %s change on table %u from %u",
+ get_dependency_type_str(key->kind),
key->relid, rientry->remote_xid);
existing_xid = rientry->remote_xid;
@@ -1007,9 +1031,8 @@ check_and_record_key_dependency(ReplicaIdentityKey *key,
if (has_active_key_dependency(rientry, false))
{
elog(DEBUG1,
- key->kind == LOGICALREP_KEY_REPLICA_IDENTITY ?
- "found conflicting replica identity change on table %u from %u" :
- "found conflicting local unique key change on table %u from %u",
+ "found conflicting %s change on table %u from %u",
+ get_dependency_type_str(key->kind),
key->relid, rientry->remote_xid);
existing_xid = rientry->remote_xid;
@@ -1039,6 +1062,23 @@ has_null_key_values(LogicalRepTupleData *data, Bitmapset *indexkeys)
return false;
}
+/*
+ * Check if any of the attnums have NULL values.
+ */
+static bool
+fk_tuple_has_null(LogicalRepTupleData *data, List *attnums)
+{
+ foreach_int(attnum, attnums)
+ {
+ int attidx = AttrNumberGetAttrOffset(attnum);
+
+ if (data->colstatus[attidx] == LOGICALREP_COLUMN_NULL)
+ return true;
+ }
+
+ return false;
+}
+
/*
* Build a hash key for replica_identity_table using the given relation and
* tuple data, restricted to the specified key columns.
@@ -1104,18 +1144,81 @@ build_replica_identity_key(Oid relid, LogicalRepTupleData *original_data,
return key;
}
+/*
+ * Similar to build_replica_identity_key(), but builds the hash key from
+ * explicitly ordered columns.
+ */
+static ReplicaIdentityKey *
+build_replica_identity_key_by_attnums(Oid relid,
+ LogicalRepTupleData *original_data,
+ List *attnums)
+{
+ LogicalRepTupleData *keydata;
+ ReplicaIdentityKey *key;
+ MemoryContext oldctx;
+ int i_key = 0;
+ int nattnums = list_length(attnums);
+
+ oldctx = MemoryContextSwitchTo(ApplyContext);
+
+ keydata = palloc0_object(LogicalRepTupleData);
+
+ if (nattnums)
+ {
+ keydata->colvalues = palloc0_array(StringInfoData, nattnums);
+ keydata->colstatus = palloc0_array(char, nattnums);
+ }
+
+ keydata->ncols = nattnums;
+
+ foreach_int(attnum, attnums)
+ {
+ int attidx = AttrNumberGetAttrOffset(attnum);
+ StringInfo original_colvalue;
+
+ keydata->colstatus[i_key] = original_data->colstatus[attidx];
+
+ if (keydata->colstatus[i_key] != LOGICALREP_COLUMN_NULL)
+ {
+ Assert(original_data->colstatus[attidx] != LOGICALREP_COLUMN_UNCHANGED ||
+ original_data->colvalues[attidx].len > 0);
+
+ original_colvalue = &original_data->colvalues[attidx];
+
+ initStringInfoExt(&keydata->colvalues[i_key], original_colvalue->len + 1);
+ appendStringInfoString(&keydata->colvalues[i_key], original_colvalue->data);
+ }
+
+ i_key++;
+ }
+
+ key = palloc0_object(ReplicaIdentityKey);
+ key->relid = relid;
+ key->data = keydata;
+
+ MemoryContextSwitchTo(oldctx);
+
+ return key;
+}
+
/*
* Build dependency key information for the given relation entry if not already
* collected.
*
- * See logicalrep_build_dependent_unique_indexes() for details.
+ * When missing_ok is true, we don't throw an error if the local relation
+ * doesn't exist.
+ *
+ * See logicalrep_build_dependent_unique_indexes() and
+ * logicalrep_build_dependent_fkeys() for details.
*/
static void
-build_local_dependent_key_info(LogicalRepRelMapEntry *relentry)
+build_local_dependent_key_info(LogicalRepRelMapEntry *relentry, bool missing_ok)
{
+ Oid localrelid = relentry->localreloid;
bool needs_start;
- if (relentry->local_unique_indexes_valid)
+ if (relentry->local_unique_indexes_valid &&
+ relentry->local_fkeys_valid)
return;
/*
@@ -1127,11 +1230,20 @@ build_local_dependent_key_info(LogicalRepRelMapEntry *relentry)
if (needs_start)
StartTransactionCommand();
- relentry = logicalrep_rel_open(relentry->remoterel.remoteid, AccessShareLock);
+ if (!OidIsValid(localrelid))
+ localrelid = RangeVarGetRelid(makeRangeVar(relentry->remoterel.nspname,
+ relentry->remoterel.relname, -1),
+ AccessShareLock, missing_ok);
- logicalrep_build_dependent_unique_indexes(relentry);
+ if (OidIsValid(localrelid))
+ {
+ relentry = logicalrep_rel_open(relentry->remoterel.remoteid, AccessShareLock);
+
+ logicalrep_build_dependent_unique_indexes(relentry);
+ logicalrep_build_dependent_fkeys(relentry);
- logicalrep_rel_close(relentry, AccessShareLock);
+ logicalrep_rel_close(relentry, AccessShareLock);
+ }
if (needs_start)
CommitTransactionCommand();
@@ -1160,7 +1272,7 @@ check_and_record_local_key_dependency(Oid relid,
Assert(relentry);
- build_local_dependent_key_info(relentry);
+ build_local_dependent_key_info(relentry, false);
foreach_ptr(LogicalRepSubUniqueIndex, idxinfo, relentry->local_unique_indexes)
{
@@ -1215,6 +1327,135 @@ check_and_record_local_key_dependency(Oid relid,
}
}
+/*
+ * Check and record dependencies caused by foreign key constraints.
+ *
+ * A new value in the referencing table depends on the referenced table's data.
+ * Conversely, a deletion in the referenced table depends on whether the
+ * referencing table still contains matching rows.
+ *
+ * See applyparallelworker.c for details on why tracking these dependencies is
+ * necessary.
+ */
+static void
+check_and_record_fkey_dependency(Oid relid,
+ LogicalRepTupleData *original_data,
+ bool old_tuple,
+ TransactionId new_depended_xid,
+ List **depends_on_xids)
+{
+ LogicalRepRelMapEntry *relentry;
+ ReplicaIdentityKey *rikey;
+
+ Assert(depends_on_xids);
+
+ relentry = logicalrep_get_relentry(relid);
+
+ Assert(relentry);
+
+ build_local_dependent_key_info(relentry, false);
+
+ foreach_ptr(LogicalRepSubFKey, fkinfo, relentry->local_fkeys)
+ {
+ List *keyattnums = old_tuple ?
+ fkinfo->fkattnums_old : fkinfo->fkattnums_new;
+
+ /* NULL value won't violate foreign key violation */
+ if (fk_tuple_has_null(original_data, keyattnums))
+ continue;
+
+ /*
+ * Old tuples of foreign keys do not conflict with any preceding
+ * transaction (see the comments in applyparallelworker.c for details on
+ * conflicting cases). When we don't need to record a new dependency, we
+ * can skip processing this index entirely.
+ */
+ if (old_tuple && !TransactionIdIsValid(new_depended_xid))
+ continue;
+
+ rikey = build_replica_identity_key_by_attnums(fkinfo->ref_remoteid,
+ original_data,
+ keyattnums);
+
+ /*
+ * For old tuples, record a dependency that later transactions deleting
+ * the same key from the referenced table must wait on. No preceding
+ * transactions are added to the list.
+ *
+ * For new tuples, check for existing dependencies on the same key
+ * inserted into the referenced table and add any dependent transactions
+ * to the list.
+ */
+ if (old_tuple)
+ {
+ rikey->kind = LOGICALREP_KEY_FOREIGN_KEY;
+ (void) check_and_record_key_dependency(rikey, new_depended_xid);
+ }
+ else
+ {
+ TransactionId dep_xid;
+
+ rikey->kind = LOGICALREP_KEY_REFERENCED_KEY;
+ dep_xid = check_and_record_key_dependency(rikey, InvalidTransactionId);
+
+ if (TransactionIdIsValid(dep_xid) &&
+ !TransactionIdEquals(dep_xid, new_depended_xid))
+ append_xid_dependency(dep_xid, depends_on_xids);
+ }
+ }
+
+ /* Return if no referenced key exists */
+ if (relentry->local_refkeys == NIL)
+ return;
+
+ foreach_ptr(LogicalRepSubRefKey, refinfo, relentry->local_refkeys)
+ {
+ List *keyattnums = old_tuple ?
+ refinfo->refattnums_old : refinfo->refattnums_new;
+
+ /* Shouldn't be real NULL value in referenced key (primary key). */
+ Assert(old_tuple || !fk_tuple_has_null(original_data, keyattnums));
+
+ /*
+ * New tuples of referenced keys do not conflict with any preceding
+ * transaction (see the comments in applyparallelworker.c for details on
+ * conflicting cases). When we don't need to record a new dependency, we
+ * can skip processing this key entirely.
+ */
+ if (!old_tuple && !TransactionIdIsValid(new_depended_xid))
+ continue;
+
+ rikey = build_replica_identity_key_by_attnums(relid, original_data,
+ keyattnums);
+
+ /*
+ * For old tuples, check for existing dependencies on the same key
+ * deleted from the referencing table and add any dependent transactions
+ * to the list.
+ *
+ * For new tuples, record a dependency that later transactions inserting
+ * the same key into the referencing table must wait on. No preceding
+ * transactions are added to the list.
+ */
+ if (old_tuple)
+ {
+ TransactionId dep_xid;
+
+ rikey->kind = LOGICALREP_KEY_FOREIGN_KEY;
+ dep_xid = check_and_record_key_dependency(rikey, InvalidTransactionId);
+
+ if (TransactionIdIsValid(dep_xid) &&
+ !TransactionIdEquals(dep_xid, new_depended_xid))
+ append_xid_dependency(dep_xid, depends_on_xids);
+ }
+ else
+ {
+ rikey->kind = LOGICALREP_KEY_REFERENCED_KEY;
+ (void) check_and_record_key_dependency(rikey, new_depended_xid);
+ }
+ }
+}
+
/*
* Check for dependencies on preceding transactions that modify the same key.
* Returns the dependent transactions in 'depends_on_xids'.
@@ -1290,6 +1531,7 @@ check_and_record_ri_dependency(Oid relid, LogicalRepTupleData *original_data,
static void
find_all_dependencies_on_rel(LogicalRepRelId relid,
TransactionId new_depended_xid,
+ LogicalRepKeyKind kind,
List **depends_on_xids)
{
replica_identity_iterator i;
@@ -1305,6 +1547,10 @@ find_all_dependencies_on_rel(LogicalRepRelId relid,
if (rientry->keydata->relid != relid)
continue;
+ if (kind != LOGICALREP_ALL_KEYS &&
+ rientry->keydata->kind != kind)
+ continue;
+
/* Skip entries that do not have a valid and uncommitted transaction */
if (!has_active_key_dependency(rientry, true))
continue;
@@ -1337,7 +1583,8 @@ check_and_record_rel_dependency(LogicalRepRelId relid,
Assert(depends_on_xids);
- find_all_dependencies_on_rel(relid, new_depended_xid, depends_on_xids);
+ find_all_dependencies_on_rel(relid, new_depended_xid, LOGICALREP_ALL_KEYS,
+ depends_on_xids);
/* Search for existing entry */
relentry = logicalrep_get_relentry(relid);
@@ -1363,6 +1610,100 @@ check_and_record_rel_dependency(LogicalRepRelId relid,
relentry->last_depended_xid = new_depended_xid;
}
+/*
+ * Check for preceding transactions that modified tables referencing or
+ * referenced by the given table. Any dependent transaction IDs are added to
+ * 'depends_on_xids'.
+ *
+ * This function is called only for TRUNCATE and remote schema change that
+ * affect the entire table.
+ *
+ * For remote schema changes, foreign key information of all referencing and
+ * referenced side tables is also invalidated.
+ */
+static void
+find_frel_dependency(LogicalRepRelId relid, TransactionId new_depended_xid,
+ LogicalRepMsgType action, List **depends_on_xids)
+{
+ LogicalRepRelMapEntry *relentry;
+ bool schema_change = (action == LOGICAL_REP_MSG_RELATION);
+
+ Assert(depends_on_xids);
+ Assert(action == LOGICAL_REP_MSG_TRUNCATE ||
+ action == LOGICAL_REP_MSG_RELATION);
+
+ /* Search for existing entry */
+ relentry = logicalrep_get_relentry(relid);
+
+ Assert(relentry);
+
+ /*
+ * The relation message does not always indicate that the corresponding
+ * local relation exists. For example, the relation message may be sent for
+ * a partition even if the changes are published as root table and the same
+ * partition may not exists on subscriber. So missing_ok is true for schema
+ * change, but false for TRUNCATE.
+ */
+ build_local_dependent_key_info(relentry, schema_change);
+
+ /*
+ * For both TRUNCATE and schema changes, wait for preceding transactions
+ * that deleted from the referencing side table.
+ */
+ foreach_ptr(LogicalRepSubRefKey, refinfo, relentry->local_refkeys)
+ {
+ find_all_dependencies_on_rel(refinfo->fk_remoteid, new_depended_xid,
+ LOGICALREP_KEY_FOREIGN_KEY,
+ depends_on_xids);
+
+ check_and_record_rel_dependency(refinfo->fk_remoteid,
+ InvalidTransactionId,
+ depends_on_xids);
+
+ if (schema_change)
+ {
+ LogicalRepRelMapEntry *fkentry;
+
+ fkentry = logicalrep_get_relentry(refinfo->fk_remoteid);
+
+ if (fkentry)
+ fkentry->local_fkeys_valid = false;
+ }
+ }
+
+ /*
+ * TRUNCATE does not need to wait for referenced-side changes, but schema
+ * changes do. A replica identity change invalidates previously recorded
+ * referenced-side dependency entries. Without waiting, a later insert on
+ * the referencing table might miss the correct dependency on the referenced
+ * table, so it's safer to wait for preceding transactions on the referenced
+ * side to finish.
+ */
+ if (!schema_change)
+ return;
+
+ foreach_ptr(LogicalRepSubFKey, fkinfo, relentry->local_fkeys)
+ {
+ find_all_dependencies_on_rel(fkinfo->ref_remoteid, new_depended_xid,
+ LOGICALREP_KEY_REFERENCED_KEY,
+ depends_on_xids);
+
+ check_and_record_rel_dependency(fkinfo->ref_remoteid,
+ InvalidTransactionId,
+ depends_on_xids);
+
+ if (schema_change)
+ {
+ LogicalRepRelMapEntry *refentry;
+
+ refentry = logicalrep_get_relentry(fkinfo->ref_remoteid);
+
+ if (refentry)
+ refentry->local_fkeys_valid = false;
+ }
+ }
+}
+
/*
* Check the parallel safety of applying the give action for the relation. If
* not safe, wait for preceding transactions to finish before proceeding with
@@ -1481,6 +1822,9 @@ handle_dependency_on_change(LogicalRepMsgType action, StringInfo s,
check_and_record_local_key_dependency(relid, &newtup, false,
new_depended_xid,
&depends_on_xids);
+ check_and_record_fkey_dependency(relid, &newtup, false,
+ new_depended_xid,
+ &depends_on_xids);
break;
case LOGICAL_REP_MSG_UPDATE:
@@ -1524,6 +1868,13 @@ handle_dependency_on_change(LogicalRepMsgType action, StringInfo s,
check_and_record_local_key_dependency(relid, &oldtup, true,
new_depended_xid,
&depends_on_xids);
+
+ check_and_record_fkey_dependency(relid, &newtup, false,
+ new_depended_xid,
+ &depends_on_xids);
+ check_and_record_fkey_dependency(relid, &oldtup, true,
+ new_depended_xid,
+ &depends_on_xids);
break;
case LOGICAL_REP_MSG_DELETE:
@@ -1533,6 +1884,9 @@ handle_dependency_on_change(LogicalRepMsgType action, StringInfo s,
check_and_record_local_key_dependency(relid, &oldtup, true,
new_depended_xid,
&depends_on_xids);
+ check_and_record_fkey_dependency(relid, &oldtup, true,
+ new_depended_xid,
+ &depends_on_xids);
break;
case LOGICAL_REP_MSG_TRUNCATE:
@@ -1545,10 +1899,15 @@ handle_dependency_on_change(LogicalRepMsgType action, StringInfo s,
* modified the same table.
*/
foreach_int(truncated_relid, remote_relids)
+ {
check_and_record_rel_dependency(truncated_relid,
new_depended_xid,
&depends_on_xids);
+ find_frel_dependency(truncated_relid, new_depended_xid,
+ action, &depends_on_xids);
+ }
+
break;
case LOGICAL_REP_MSG_RELATION:
@@ -1562,6 +1921,12 @@ handle_dependency_on_change(LogicalRepMsgType action, StringInfo s,
*/
check_and_record_rel_dependency(rel->remoteid, new_depended_xid,
&depends_on_xids);
+
+ logicalrep_relmap_update(rel);
+
+ find_frel_dependency(rel->remoteid, new_depended_xid, action,
+ &depends_on_xids);
+
break;
case LOGICAL_REP_MSG_TYPE:
diff --git a/src/include/replication/logicalrelation.h b/src/include/replication/logicalrelation.h
index ec4c9ece571..a917ce2fe86 100644
--- a/src/include/replication/logicalrelation.h
+++ b/src/include/replication/logicalrelation.h
@@ -60,6 +60,11 @@ typedef struct LogicalRepRelMapEntry
List *local_unique_indexes;
bool local_unique_indexes_valid;
+ /* Local foreign keys. Used for dependency tracking */
+ List *local_fkeys;
+ List *local_refkeys;
+ bool local_fkeys_valid;
+
/*
* Per-operation safety cache for parallel apply. If
* parallel_global_unsafe[action] is true, that action cannot be applied in
@@ -89,6 +94,30 @@ typedef struct LogicalRepSubUniqueIndex
bool nulls_distinct; /* Whether NULLs are considered distinct */
} LogicalRepSubUniqueIndex;
+/*
+ * Referencing side foreign key information. This is used to track dependencies
+ * between transactions that modify the same referenced and referencing key
+ * values.
+ */
+typedef struct LogicalRepSubFKey
+{
+ Oid conoid; /* OID of the FK constraint */
+ LogicalRepRelId ref_remoteid; /* referenced remote relation */
+ List *fkattnums_new; /* new-tuple-safe referencing remote attnums */
+ List *fkattnums_old; /* old-tuple-safe referencing remote attnums */
+} LogicalRepSubFKey;
+
+/*
+ * Similar to LogicalRepSubFKey, but for referenced-side primary keys.
+ */
+typedef struct LogicalRepSubRefKey
+{
+ Oid conoid; /* OID of the FK constraint */
+ LogicalRepRelId fk_remoteid; /* referencing remote relation */
+ List *refattnums_new; /* new-tuple-safe referenced remote attnums */
+ List *refattnums_old; /* old-tuple-safe referenced remote attnums */
+} LogicalRepSubRefKey;
+
extern void logicalrep_relmap_update(LogicalRepRelation *remoterel);
extern void logicalrep_partmap_reset_relmap(LogicalRepRelation *remoterel);
@@ -100,6 +129,7 @@ extern void logicalrep_rel_close(LogicalRepRelMapEntry *rel,
LOCKMODE lockmode);
extern void logicalrep_rel_check_parallel_safety(LogicalRepRelMapEntry *entry);
extern void logicalrep_build_dependent_unique_indexes(LogicalRepRelMapEntry *entry);
+extern void logicalrep_build_dependent_fkeys(LogicalRepRelMapEntry *entry);
extern bool IsIndexUsableForReplicaIdentityFull(Relation idxrel, AttrMap *attrmap);
extern Oid GetRelationIdentityOrPK(Relation rel);
extern int logicalrep_get_num_rels(void);
diff --git a/src/test/subscription/t/050_parallel_apply.pl b/src/test/subscription/t/050_parallel_apply.pl
index 9e981bab981..9d52cf660f8 100644
--- a/src/test/subscription/t/050_parallel_apply.pl
+++ b/src/test/subscription/t/050_parallel_apply.pl
@@ -37,6 +37,35 @@ $node_publisher->safe_psql(
CREATE UNIQUE INDEX tab_toast_ri_index on tab_toast (a, b);
ALTER TABLE tab_toast REPLICA IDENTITY USING INDEX tab_toast_ri_index;
INSERT INTO tab_toast(a, b) VALUES(repeat('1234567890', 200), '1234567890');
+
+ CREATE TABLE regress_tab_pk (id int PRIMARY KEY);
+ CREATE TABLE regress_tab_fk (id int PRIMARY KEY, fk int REFERENCES regress_tab_pk (id));
+
+ CREATE TABLE pk_parted (
+ id int PRIMARY KEY,
+ value int
+ ) PARTITION BY RANGE (id);
+
+ CREATE TABLE pk_parted_1 (
+ value int,
+ id int NOT NULL
+ );
+
+ ALTER TABLE pk_parted ATTACH PARTITION pk_parted_1
+ FOR VALUES FROM (1) TO (10);
+
+ CREATE TABLE fk_parted (
+ id int REFERENCES pk_parted(id),
+ value int
+ ) PARTITION BY RANGE (id);
+
+ CREATE TABLE fk_parted_1 (
+ value int,
+ id int
+ );
+
+ ALTER TABLE fk_parted ATTACH PARTITION fk_parted_1
+ FOR VALUES FROM (1) TO (10);
));
$node_publisher->safe_psql('postgres',
"INSERT INTO regress_tab VALUES (generate_series(1, 10), 'test');");
@@ -78,6 +107,35 @@ $node_subscriber->safe_psql(
ALTER TABLE tab_toast ALTER COLUMN a SET STORAGE EXTERNAL;
CREATE UNIQUE INDEX tab_toast_ri_index on tab_toast (a, b);
ALTER TABLE tab_toast REPLICA IDENTITY USING INDEX tab_toast_ri_index;
+
+ CREATE TABLE regress_tab_pk (id int PRIMARY KEY);
+ CREATE TABLE regress_tab_fk (id int PRIMARY KEY, fk int REFERENCES regress_tab_pk (id));
+
+ CREATE TABLE pk_parted (
+ id int PRIMARY KEY,
+ value int
+ ) PARTITION BY RANGE (id);
+
+ CREATE TABLE pk_parted_1 (
+ value int,
+ id int NOT NULL
+ );
+
+ ALTER TABLE pk_parted ATTACH PARTITION pk_parted_1
+ FOR VALUES FROM (1) TO (10);
+
+ CREATE TABLE fk_parted (
+ id int REFERENCES pk_parted(id),
+ value int
+ ) PARTITION BY RANGE (id);
+
+ CREATE TABLE fk_parted_1 (
+ value int,
+ id int
+ );
+
+ ALTER TABLE fk_parted ATTACH PARTITION fk_parted_1
+ FOR VALUES FROM (1) TO (10);
));
$node_subscriber->safe_psql('postgres',
"CREATE SUBSCRIPTION regress_sub CONNECTION '$publisher_connstr' PUBLICATION regress_pub;");
@@ -651,4 +709,276 @@ $node_subscriber->safe_psql('postgres', "DROP INDEX local_unique_idx_expr;");
$node_publisher->safe_psql('postgres', "TRUNCATE TABLE regress_tab;");
$node_publisher->wait_for_catchup('regress_sub');
+##################################################
+# Test that the dependency tracking works correctly for foreign keys on
+# subscriber during parallel apply.
+##################################################
+
+# Test that when receiving table schema information for a referencing table, the
+# subscriber correctly checks for dependencies on the referenced table and waits
+# for its changes to complete.
+$node_subscriber->safe_psql('postgres',
+ "SELECT injection_points_attach('parallel-worker-before-commit','wait');"
+);
+
+# Enable foreign triggers on subscriber
+my $pk_fk_trigger = $node_subscriber->safe_psql('postgres', qq[
+ SELECT tgname
+ FROM pg_trigger
+ WHERE tgrelid = 'regress_tab_pk'::regclass
+ AND tgconstraint > 0
+ LIMIT 1
+]);
+chomp($pk_fk_trigger);
+
+my $fk_fk_trigger = $node_subscriber->safe_psql('postgres', qq[
+ SELECT tgname
+ FROM pg_trigger
+ WHERE tgrelid = 'regress_tab_fk'::regclass
+ AND tgconstraint > 0
+ LIMIT 1
+]);
+chomp($fk_fk_trigger);
+
+$node_subscriber->safe_psql('postgres',
+ qq[ALTER TABLE regress_tab_pk ENABLE REPLICA TRIGGER "$pk_fk_trigger";
+ ALTER TABLE regress_tab_fk ENABLE REPLICA TRIGGER "$fk_fk_trigger";]);
+
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO regress_tab_pk VALUES (2);");
+
+$node_subscriber->wait_for_event('logical replication parallel worker',
+ 'parallel-worker-before-commit');
+
+$offset = -s $node_subscriber->logfile;
+
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO regress_tab_fk VALUES (1, 2);");
+
+# Verify the dependency is detected on referenced table schema information
+$str = $node_subscriber->wait_for_log(qr/found conflicting change on table [1-9][0-9]+ from ([1-9][0-9]+)/, $offset);
+$xid = $str =~ /found conflicting change on table [1-9][0-9]+ from ([1-9][0-9]+)/;
+
+# Verify the parallel worker waits for the same transaction
+$node_subscriber->wait_for_log(qr/wait for depended xid $xid/, $offset);
+
+ok(1, "referenced table dependency detected for parallel apply");
+
+# Wakeup the parallel worker
+$node_subscriber->safe_psql('postgres', qq[
+ SELECT injection_points_detach('parallel-worker-before-commit');
+ SELECT injection_points_wakeup('parallel-worker-before-commit');
+]);
+
+# Verify the streamed transaction can be applied
+$node_subscriber->wait_for_log(qr/finish waiting for depended xid $xid/, $offset);
+$node_publisher->wait_for_catchup('regress_sub');
+
+$result =
+ $node_subscriber->safe_psql('postgres', "SELECT count(1) FROM regress_tab_pk");
+is ($result, 1, 'insert is replicated to referenced table on subscriber');
+
+$result =
+ $node_subscriber->safe_psql('postgres', "SELECT count(1) FROM regress_tab_fk");
+is ($result, 1, 'insert is replicated to referencing table on subscriber');
+
+# INSERT - INSERT case: Tx-1 inserts referenced tuple and Tx-2 inserts
+# referencing tuple.
+
+$node_subscriber->safe_psql('postgres',
+ "SELECT injection_points_attach('parallel-worker-before-commit','wait');"
+);
+
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO regress_tab_pk VALUES (3);");
+
+$node_subscriber->wait_for_event('logical replication parallel worker',
+ 'parallel-worker-before-commit');
+
+$offset = -s $node_subscriber->logfile;
+
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO regress_tab_fk VALUES (2, 3);");
+
+# Verify the dependency for referenced key change is detected
+$str = $node_subscriber->wait_for_log(qr/found conflicting referenced key change on table [1-9][0-9]+ from ([1-9][0-9]+)/, $offset);
+$xid = $str =~ /found conflicting referenced key change on table [1-9][0-9]+ from ([1-9][0-9]+)/;
+
+# Verify the parallel worker waits for the same transaction
+$node_subscriber->wait_for_log(qr/wait for depended xid $xid/, $offset);
+
+ok(1, "referenced key dependency detected for parallel apply");
+
+# Wakeup the parallel worker
+$node_subscriber->safe_psql('postgres', qq[
+ SELECT injection_points_detach('parallel-worker-before-commit');
+ SELECT injection_points_wakeup('parallel-worker-before-commit');
+]);
+
+# Verify the streamed transaction can be applied
+$node_subscriber->wait_for_log(qr/finish waiting for depended xid $xid/, $offset);
+$node_publisher->wait_for_catchup('regress_sub');
+
+$result =
+ $node_subscriber->safe_psql('postgres', "SELECT count(1) FROM regress_tab_pk");
+is ($result, 2, 'insert is replicated to referenced table on subscriber');
+
+$result =
+ $node_subscriber->safe_psql('postgres', "SELECT count(1) FROM regress_tab_fk");
+is ($result, 2, 'insert is replicated to referencing table on subscriber');
+
+# DELETE - DELETE case: Tx-1 deletes referencing tuple and Tx-2 deletes
+# referenced tuple.
+$node_subscriber->safe_psql('postgres',
+ "SELECT injection_points_attach('parallel-worker-before-commit','wait');"
+);
+
+$node_publisher->safe_psql('postgres',
+ "DELETE FROM regress_tab_fk WHERE id = 1;");
+
+$node_subscriber->wait_for_event('logical replication parallel worker',
+ 'parallel-worker-before-commit');
+
+$offset = -s $node_subscriber->logfile;
+
+$node_publisher->safe_psql('postgres',
+ "DELETE FROM regress_tab_pk WHERE id = 2;");
+
+$str = $node_subscriber->wait_for_log(qr/found conflicting foreign key change on table [1-9][0-9]+ from ([1-9][0-9]+)/, $offset);
+$xid = $str =~ /found conflicting foreign key change on table [1-9][0-9]+ from ([1-9][0-9]+)/;
+
+$node_subscriber->wait_for_log(qr/wait for depended xid $xid/, $offset);
+
+ok(1, "foreign key dependency detected for parallel apply");
+
+$node_subscriber->safe_psql('postgres', qq[
+ SELECT injection_points_detach('parallel-worker-before-commit');
+ SELECT injection_points_wakeup('parallel-worker-before-commit');
+]);
+
+$node_subscriber->wait_for_log(qr/finish waiting for depended xid $xid/, $offset);
+$node_publisher->wait_for_catchup('regress_sub');
+
+$result =
+ $node_subscriber->safe_psql('postgres', "SELECT count(1) FROM regress_tab_pk");
+is ($result, 1, 'delete is replicated to referenced table on subscriber');
+
+$result =
+ $node_subscriber->safe_psql('postgres', "SELECT count(1) FROM regress_tab_fk");
+is ($result, 1, 'delete is replicated to referencing table on subscriber');
+
+##################################################
+# Test that the dependency tracking works correctly for foreign keys when both
+# the referenced and referencing tables are partitioned.
+##################################################
+
+$node_subscriber->safe_psql('postgres',
+ "SELECT injection_points_attach('parallel-worker-before-commit','wait');"
+);
+
+my $part_pk_trigger = $node_subscriber->safe_psql('postgres', qq[
+ SELECT tgname
+ FROM pg_trigger
+ WHERE tgrelid = 'pk_parted_1'::regclass
+ AND tgconstraint > 0
+ LIMIT 1
+]);
+chomp($part_pk_trigger);
+
+my $part_fk_trigger = $node_subscriber->safe_psql('postgres', qq[
+ SELECT tgname
+ FROM pg_trigger
+ WHERE tgrelid = 'fk_parted_1'::regclass
+ AND tgconstraint > 0
+ LIMIT 1
+]);
+chomp($part_fk_trigger);
+
+$node_subscriber->safe_psql('postgres',
+ qq[ALTER TABLE pk_parted_1 ENABLE REPLICA TRIGGER "$part_pk_trigger";
+ ALTER TABLE fk_parted_1 ENABLE REPLICA TRIGGER "$part_fk_trigger";]);
+
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO pk_parted_1(id, value) VALUES (1, 10);");
+
+$node_subscriber->wait_for_event('logical replication parallel worker',
+ 'parallel-worker-before-commit');
+
+$offset = -s $node_subscriber->logfile;
+
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO fk_parted_1(id, value) VALUES (1, 10);");
+
+# Verify the dependency is detected on referenced table schema information
+$str = $node_subscriber->wait_for_log(qr/found conflicting change on table [1-9][0-9]+ from ([1-9][0-9]+)/, $offset);
+$xid = $str =~ /found conflicting change on table [1-9][0-9]+ from ([1-9][0-9]+)/;
+
+# Verify the parallel worker waits for the same transaction
+$node_subscriber->wait_for_log(qr/wait for depended xid $xid/, $offset);
+
+ok(1, "partitioned referenced table dependency detected for parallel apply");
+
+# Wakeup the parallel worker
+$node_subscriber->safe_psql('postgres', qq[
+ SELECT injection_points_detach('parallel-worker-before-commit');
+ SELECT injection_points_wakeup('parallel-worker-before-commit');
+]);
+
+# Verify the streamed transaction can be applied
+$node_subscriber->wait_for_log(qr/finish waiting for depended xid $xid/, $offset);
+$node_publisher->wait_for_catchup('regress_sub');
+
+$result =
+ $node_subscriber->safe_psql('postgres', "SELECT count(1) FROM pk_parted_1");
+is ($result, 1, 'insert is replicated to referenced table on subscriber');
+
+$result =
+ $node_subscriber->safe_psql('postgres', "SELECT count(1) FROM fk_parted_1");
+is ($result, 1, 'insert is replicated to referencing table on subscriber');
+
+# INSERT - INSERT case: Tx-1 inserts referenced tuple and Tx-2 inserts
+# referencing tuple.
+
+$node_subscriber->safe_psql('postgres',
+ "SELECT injection_points_attach('parallel-worker-before-commit','wait');"
+);
+
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO pk_parted_1(id, value) VALUES (2, 20);");
+
+$node_subscriber->wait_for_event('logical replication parallel worker',
+ 'parallel-worker-before-commit');
+
+$offset = -s $node_subscriber->logfile;
+
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO fk_parted_1(id, value) VALUES (2, 20);");
+
+# Verify the dependency for referenced key change is detected
+$str = $node_subscriber->wait_for_log(qr/found conflicting referenced key change on table [1-9][0-9]+ from ([1-9][0-9]+)/, $offset);
+$xid = $str =~ /found conflicting referenced key change on table [1-9][0-9]+ from ([1-9][0-9]+)/;
+
+# Verify the parallel worker waits for the same transaction
+$node_subscriber->wait_for_log(qr/wait for depended xid $xid/, $offset);
+
+ok(1, "partitioned referenced key dependency detected for parallel apply");
+
+# Wakeup the parallel worker
+$node_subscriber->safe_psql('postgres', qq[
+ SELECT injection_points_detach('parallel-worker-before-commit');
+ SELECT injection_points_wakeup('parallel-worker-before-commit');
+]);
+
+# Verify the streamed transaction can be applied
+$node_subscriber->wait_for_log(qr/finish waiting for depended xid $xid/, $offset);
+$node_publisher->wait_for_catchup('regress_sub');
+
+$result =
+ $node_subscriber->safe_psql('postgres', "SELECT count(1) FROM pk_parted_1");
+is ($result, 2, 'insert is replicated to referenced table on subscriber');
+
+$result =
+ $node_subscriber->safe_psql('postgres', "SELECT count(1) FROM fk_parted_1");
+is ($result, 2, 'insert is replicated to referencing table on subscriber');
+
done_testing();
--
2.43.0
[application/octet-stream] v21-0007-Handle-unsafe-parallel-apply-cases-due-to-non-im.patch (20.1K, ../TY4PR01MB177180394D8E9C6D263E117E494FE2@TY4PR01MB17718.jpnprd01.prod.outlook.com/4-v21-0007-Handle-unsafe-parallel-apply-cases-due-to-non-im.patch)
download | inline diff:
From 28cff2370f49d34ca742254656e778dbef1e1fb5 Mon Sep 17 00:00:00 2001
From: Zhijie Hou <[email protected]>
Date: Tue, 9 Jun 2026 18:45:02 +0800
Subject: [PATCH v21 7/9] Handle unsafe parallel apply cases due to
non-immutable functions
This patch adds table-level parallel safety checks that mark operations as
unsafe when a table has triggers executing non-immutable functions, or when
subscriber-only column defaults contain non-immutable functions.
Safety is determined by the worker that actually applies the changes, rather
than solely by the leader. Leader-only evaluation is unsafe because the table
could be altered before changes are dispatched to a parallel worker. Without
re-validation by the parallel worker, it might incorrectly assume the relation
is safe for parallel apply.
When an unsafe operation is encountered, the worker waits for the last
parallelized transaction to commit before proceeding.
---
.../replication/logical/applyparallelworker.c | 7 +-
src/backend/replication/logical/relation.c | 98 +++++++++++++++++++
src/backend/replication/logical/worker.c | 85 ++++++++++++++--
src/include/replication/logicalrelation.h | 28 ++++++
src/include/replication/worker_internal.h | 6 +-
src/test/subscription/t/050_parallel_apply.pl | 71 ++++++++++++++
6 files changed, 286 insertions(+), 9 deletions(-)
diff --git a/src/backend/replication/logical/applyparallelworker.c b/src/backend/replication/logical/applyparallelworker.c
index 6e478a41e73..3e436b518bd 100644
--- a/src/backend/replication/logical/applyparallelworker.c
+++ b/src/backend/replication/logical/applyparallelworker.c
@@ -667,13 +667,17 @@ pa_launch_parallel_worker(void)
/*
* Allocate a parallel apply worker that will be used for the specified xid.
*
+ * preceding_xid is the transaction received before the current one; pass
+ * InvalidTransactionId if none exists.
+ *
* We first try to get an available worker from the pool, if any and then try
* to launch a new worker. On successful allocation, remember the worker
* information in the hash table so that we can get it later for processing the
* streaming changes.
*/
void
-pa_allocate_worker(TransactionId xid, bool stream_txn)
+pa_allocate_worker(TransactionId xid, TransactionId preceding_xid,
+ bool stream_txn)
{
bool found;
ParallelApplyWorkerInfo *winfo = NULL;
@@ -710,6 +714,7 @@ pa_allocate_worker(TransactionId xid, bool stream_txn)
SpinLockAcquire(&winfo->shared->mutex);
winfo->shared->xact_state = PARALLEL_TRANS_UNKNOWN;
winfo->shared->xid = xid;
+ winfo->shared->preceding_xid = preceding_xid;
SpinLockRelease(&winfo->shared->mutex);
winfo->in_use = true;
diff --git a/src/backend/replication/logical/relation.c b/src/backend/replication/logical/relation.c
index c5a83b095fa..6d3e1827ae2 100644
--- a/src/backend/replication/logical/relation.c
+++ b/src/backend/replication/logical/relation.c
@@ -21,11 +21,16 @@
#include "access/genam.h"
#include "access/table.h"
#include "catalog/namespace.h"
+#include "catalog/pg_proc.h"
#include "catalog/pg_subscription_rel.h"
+#include "catalog/pg_trigger.h"
+#include "commands/trigger.h"
#include "executor/executor.h"
#include "nodes/makefuncs.h"
+#include "optimizer/optimizer.h"
#include "replication/logicalrelation.h"
#include "replication/worker_internal.h"
+#include "rewrite/rewriteHandler.h"
#include "utils/inval.h"
#include "utils/lsyscache.h"
#include "utils/syscache.h"
@@ -82,6 +87,7 @@ logicalrep_relmap_invalidate_cb(Datum arg, Oid reloid)
if (entry->localreloid == reloid)
{
entry->localrelvalid = false;
+ entry->parallel_safety_valid = false;
hash_seq_term(&status);
break;
}
@@ -95,7 +101,10 @@ logicalrep_relmap_invalidate_cb(Datum arg, Oid reloid)
hash_seq_init(&status, LogicalRepRelMap);
while ((entry = (LogicalRepRelMapEntry *) hash_seq_search(&status)) != NULL)
+ {
entry->localrelvalid = false;
+ entry->parallel_safety_valid = false;
+ }
}
}
@@ -394,6 +403,7 @@ logicalrep_rel_open(LogicalRepRelId remoteid, LOCKMODE lockmode)
{
/* Table was renamed or dropped. */
entry->localrelvalid = false;
+ entry->parallel_safety_valid = false;
}
else if (!entry->localrelvalid)
{
@@ -520,6 +530,90 @@ logicalrep_rel_close(LogicalRepRelMapEntry *rel, LOCKMODE lockmode)
rel->localrel = NULL;
}
+/*
+ * Check whether changes on this relation can be applied in parallel.
+ *
+ * Parallel apply is unsafe if the table has any user-defined functions that may
+ * be executed during change application (e.g., in triggers, defaults, or
+ * constraint expressions). In theory, we should check the volatility of all
+ * such functions.
+ *
+ * Note that we do not check user-defined CHECK constraints here, as PostgreSQL
+ * already assumes they are immutable; we follow that same rule.
+ *
+ * If any mutable function is found, the relation is marked as globally unsafe,
+ * and the result is cached.
+ */
+void
+logicalrep_rel_check_parallel_safety(LogicalRepRelMapEntry *entry)
+{
+ Relation localrel = entry->localrel;
+ TriggerDesc *trigdesc = localrel->trigdesc;
+ int ntriggers = trigdesc ? trigdesc->numtriggers : 0;
+ TupleDesc desc = RelationGetDescr(localrel);
+
+ if (entry->parallel_safety_valid)
+ return;
+
+ memset(entry->parallel_global_unsafe, 0,
+ sizeof(entry->parallel_global_unsafe));
+
+ /* Seek triggers one by one to see the volatility */
+ for (int i = 0; i < ntriggers; i++)
+ {
+ Trigger *trigger = &trigdesc->triggers[i];
+
+ Assert(OidIsValid(trigger->tgfoid));
+
+ /* Skip if the trigger is internal (e.g., fkey triggers) */
+ if (trigger->tgisinternal)
+ continue;
+
+ /* Skip if the trigger is not enabled for logical replication */
+ if (trigger->tgenabled == TRIGGER_DISABLED ||
+ trigger->tgenabled == TRIGGER_FIRES_ON_ORIGIN)
+ continue;
+
+ /* Check the volatility of the trigger. Exit if it is not immutable */
+ if (func_volatile(trigger->tgfoid) != PROVOLATILE_IMMUTABLE)
+ {
+ if (TRIGGER_FOR_INSERT(trigger->tgtype))
+ entry->parallel_global_unsafe[LRPA_INSERT] = true;
+ else if (TRIGGER_FOR_UPDATE(trigger->tgtype))
+ entry->parallel_global_unsafe[LRPA_UPDATE] = true;
+ else if (TRIGGER_FOR_DELETE(trigger->tgtype))
+ entry->parallel_global_unsafe[LRPA_DELETE] = true;
+ else if (TRIGGER_FOR_TRUNCATE(trigger->tgtype))
+ entry->parallel_global_unsafe[LRPA_TRUNCATE] = true;
+ }
+ }
+
+ /* Check column defaults for mutable functions */
+ for (int attnum = 0; attnum < desc->natts; attnum++)
+ {
+ CompactAttribute *cattr = TupleDescCompactAttr(desc, attnum);
+ Expr *defexpr;
+
+ if (cattr->attisdropped || cattr->attgenerated)
+ continue;
+
+ /* Skip columns whose values come from remote change */
+ if (entry->attrmap->attnums[attnum] >= 0)
+ continue;
+
+ defexpr = (Expr *) build_column_default(entry->localrel, attnum + 1);
+
+ if (contain_mutable_functions_after_planning(defexpr))
+ {
+ /* Only INSERT operations are affected by column defaults */
+ entry->parallel_global_unsafe[LRPA_INSERT] = true;
+ break;
+ }
+ }
+
+ entry->parallel_safety_valid = true;
+}
+
/*
* Partition cache: look up partition LogicalRepRelMapEntry's
*
@@ -553,6 +647,7 @@ logicalrep_partmap_invalidate_cb(Datum arg, Oid reloid)
if (entry->relmapentry.localreloid == reloid)
{
entry->relmapentry.localrelvalid = false;
+ entry->relmapentry.parallel_safety_valid = false;
hash_seq_term(&status);
break;
}
@@ -566,7 +661,10 @@ logicalrep_partmap_invalidate_cb(Datum arg, Oid reloid)
hash_seq_init(&status, LogicalRepPartMap);
while ((entry = (LogicalRepPartMapEntry *) hash_seq_search(&status)) != NULL)
+ {
entry->relmapentry.localrelvalid = false;
+ entry->relmapentry.parallel_safety_valid = false;
+ }
}
}
diff --git a/src/backend/replication/logical/worker.c b/src/backend/replication/logical/worker.c
index 0afba345713..18d71ceabe6 100644
--- a/src/backend/replication/logical/worker.c
+++ b/src/backend/replication/logical/worker.c
@@ -1193,6 +1193,54 @@ check_and_record_rel_dependency(LogicalRepRelId relid,
relentry->last_depended_xid = new_depended_xid;
}
+/*
+ * Check the parallel safety of applying the give action for the relation. If
+ * not safe, wait for preceding transactions to finish before proceeding with
+ * the current change.
+ *
+ * See logicalrep_rel_check_parallel_safety for details on how the safety is
+ * determined.
+ */
+static void
+check_relation_parallel_apply_safety(LogicalRepRelMapEntry *relentry,
+ LogicalRepParallelAction action)
+{
+ /* Do not check parallel apply safety for streamed transactions */
+ if (in_streamed_transaction)
+ return;
+
+ /* Parallel apply only involves the leader and parallel apply workers */
+ if (!am_leader_apply_worker() && !am_parallel_apply_worker())
+ return;
+
+ /*
+ * For partitioned tables, we only need to care if the target partition is
+ * parallel apply safe or not.
+ */
+ if (relentry->localrel->rd_rel->relkind == RELKIND_PARTITIONED_TABLE)
+ return;
+
+ logicalrep_rel_check_parallel_safety(relentry);
+
+ /* Return if the action is safe for parallel apply */
+ if (!relentry->parallel_global_unsafe[action])
+ return;
+
+ elog(DEBUG1, "found parallel unsafe change on table %u for action %d",
+ relentry->remoterel.remoteid, action);
+
+ /*
+ * Wait for preceding transactions to finish before proceeding and reset the
+ * preceding_xid to InvalidTransactionId to avoid waiting for the same
+ * transaction again in the next change.
+ */
+ if (TransactionIdIsValid(last_parallelized_remote_xid))
+ {
+ pa_wait_for_depended_transaction(last_parallelized_remote_xid);
+ last_parallelized_remote_xid = InvalidTransactionId;
+ }
+}
+
/*
* Check dependencies related to the current change by determining if the
* modification impacts the same row or table as another ongoing transaction.
@@ -2048,7 +2096,7 @@ apply_handle_begin(StringInfo s)
maybe_start_skipping_changes(begin_data.final_lsn);
- pa_allocate_worker(remote_xid, false);
+ pa_allocate_worker(remote_xid, last_parallelized_remote_xid, false);
apply_action = get_transaction_apply_action(remote_xid, &winfo);
@@ -2080,6 +2128,8 @@ apply_handle_begin(StringInfo s)
/* Hold the lock until the end of the transaction. */
pa_lock_transaction(MyParallelShared->xid, AccessExclusiveLock);
pa_set_xact_state(MyParallelShared, PARALLEL_TRANS_STARTED);
+
+ last_parallelized_remote_xid = MyParallelShared->preceding_xid;
break;
default:
@@ -2232,6 +2282,8 @@ apply_handle_commit(StringInfo s)
pa_commit_transaction();
pa_unlock_transaction(remote_xid, AccessExclusiveLock);
+
+ last_parallelized_remote_xid = InvalidTransactionId;
break;
default:
@@ -2283,7 +2335,7 @@ apply_handle_begin_prepare(StringInfo s)
maybe_start_skipping_changes(begin_data.prepare_lsn);
- pa_allocate_worker(remote_xid, false);
+ pa_allocate_worker(remote_xid, last_parallelized_remote_xid, false);
apply_action = get_transaction_apply_action(remote_xid, &winfo);
@@ -2500,6 +2552,8 @@ apply_handle_prepare(StringInfo s)
pa_unlock_transaction(MyParallelShared->xid, AccessExclusiveLock);
pa_reset_subtrans();
+
+ last_parallelized_remote_xid = InvalidTransactionId;
break;
default:
@@ -2926,7 +2980,7 @@ apply_handle_stream_start(StringInfo s)
/* Try to allocate a worker for the streaming transaction. */
if (first_segment)
- pa_allocate_worker(stream_xid, true);
+ pa_allocate_worker(stream_xid, last_parallelized_remote_xid, true);
apply_action = get_transaction_apply_action(stream_xid, &winfo);
@@ -3873,6 +3927,9 @@ apply_handle_insert(StringInfo s)
/* Set relation for error callback */
apply_error_callback_arg.rel = rel;
+ /* Check if the relation is safe for parallel apply */
+ check_relation_parallel_apply_safety(rel, LRPA_INSERT);
+
/* Initialize the executor state. */
edata = create_edata_for_relation(rel);
estate = edata->estate;
@@ -4030,6 +4087,9 @@ apply_handle_update(StringInfo s)
/* Check if we can do the update. */
check_relation_updatable(rel);
+ /* Check if the relation is safe for parallel apply */
+ check_relation_parallel_apply_safety(rel, LRPA_UPDATE);
+
/*
* Make sure that any user-supplied code runs as the table owner, unless
* the user has opted out of that behavior.
@@ -4254,6 +4314,9 @@ apply_handle_delete(StringInfo s)
/* Check if we can do the delete. */
check_relation_updatable(rel);
+ /* Check if the relation is safe for parallel apply */
+ check_relation_parallel_apply_safety(rel, LRPA_DELETE);
+
/*
* Make sure that any user-supplied code runs as the table owner, unless
* the user has opted out of that behavior.
@@ -4623,22 +4686,25 @@ apply_handle_tuple_routing(ApplyExecutionData *edata,
}
MemoryContextSwitchTo(oldctx);
+ part_entry = logicalrep_partition_open(relmapentry, partrel,
+ attrmap);
+
/* Check if we can do the update or delete on the leaf partition. */
if (operation == CMD_UPDATE || operation == CMD_DELETE)
- {
- part_entry = logicalrep_partition_open(relmapentry, partrel,
- attrmap);
check_relation_updatable(part_entry);
- }
switch (operation)
{
case CMD_INSERT:
+ check_relation_parallel_apply_safety(part_entry,
+ LRPA_INSERT);
apply_handle_insert_internal(edata, partrelinfo,
remoteslot_part);
break;
case CMD_DELETE:
+ check_relation_parallel_apply_safety(part_entry,
+ LRPA_DELETE);
apply_handle_delete_internal(edata, partrelinfo,
remoteslot_part,
part_entry->localindexoid);
@@ -4660,6 +4726,9 @@ apply_handle_tuple_routing(ApplyExecutionData *edata,
EPQState epqstate;
ConflictTupleInfo conflicttuple = {0};
+ check_relation_parallel_apply_safety(part_entry,
+ LRPA_UPDATE);
+
/* Get the matching local tuple from the partition. */
found = FindReplTupleInLocalRel(edata, partrel,
&part_entry->remoterel,
@@ -4890,6 +4959,8 @@ apply_handle_truncate(StringInfo s)
continue;
}
+ check_relation_parallel_apply_safety(rel, LRPA_TRUNCATE);
+
remote_rels = lappend(remote_rels, rel);
TargetPrivilegesCheck(rel->localrel, ACL_TRUNCATE);
rels = lappend(rels, rel->localrel);
diff --git a/src/include/replication/logicalrelation.h b/src/include/replication/logicalrelation.h
index cd852337165..439243be742 100644
--- a/src/include/replication/logicalrelation.h
+++ b/src/include/replication/logicalrelation.h
@@ -16,6 +16,16 @@
#include "catalog/index.h"
#include "replication/logicalproto.h"
+typedef enum LogicalRepParallelAction
+{
+ LRPA_INSERT,
+ LRPA_UPDATE,
+ LRPA_DELETE,
+ LRPA_TRUNCATE
+} LogicalRepParallelAction;
+
+#define LRPA_ACTION_COUNT (LRPA_TRUNCATE + 1)
+
typedef struct LogicalRepRelMapEntry
{
LogicalRepRelation remoterel; /* key is remoterel.remoteid */
@@ -45,6 +55,23 @@ typedef struct LogicalRepRelMapEntry
* applying (see check_dependency_on_rel).
*/
TransactionId last_depended_xid;
+
+ /*
+ * Per-operation safety cache for parallel apply. If
+ * parallel_global_unsafe[action] is true, that action cannot be applied in
+ * parallel for this relation.
+ *
+ * This cache must be computed by the worker that actually applies changes
+ * (via logicalrep_rel_check_parallel_safety), rather than solely by the
+ * leader.
+ *
+ * Relying on the leader alone is unsafe because the table could be altered
+ * before changes are dispatched to a parallel worker. Without re-validation
+ * by the parallel worker, it might incorrectly assume that parallel apply
+ * is safe for this relation.
+ */
+ bool parallel_safety_valid;
+ bool parallel_global_unsafe[LRPA_ACTION_COUNT];
} LogicalRepRelMapEntry;
extern void logicalrep_relmap_update(LogicalRepRelation *remoterel);
@@ -56,6 +83,7 @@ extern LogicalRepRelMapEntry *logicalrep_partition_open(LogicalRepRelMapEntry *r
Relation partrel, AttrMap *map);
extern void logicalrep_rel_close(LogicalRepRelMapEntry *rel,
LOCKMODE lockmode);
+extern void logicalrep_rel_check_parallel_safety(LogicalRepRelMapEntry *entry);
extern bool IsIndexUsableForReplicaIdentityFull(Relation idxrel, AttrMap *attrmap);
extern Oid GetRelationIdentityOrPK(Relation rel);
extern int logicalrep_get_num_rels(void);
diff --git a/src/include/replication/worker_internal.h b/src/include/replication/worker_internal.h
index 3dde899f465..4f6d699492f 100644
--- a/src/include/replication/worker_internal.h
+++ b/src/include/replication/worker_internal.h
@@ -205,6 +205,9 @@ typedef struct ParallelApplyWorkerShared
*/
dsa_handle parallel_apply_dsa_handle;
dshash_table_handle parallelized_txns_handle;
+
+ /* The remote transaction ID received before the current transaction */
+ TransactionId preceding_xid;
} ParallelApplyWorkerShared;
/*
@@ -353,7 +356,8 @@ extern void apply_error_callback(void *arg);
extern void set_apply_error_context_origin(char *originname);
/* Parallel apply worker setup and interactions */
-extern void pa_allocate_worker(TransactionId xid, bool stream_txn);
+extern void pa_allocate_worker(TransactionId xid, TransactionId preceding_xid,
+ bool stream_txn);
extern ParallelApplyWorkerInfo *pa_find_worker(TransactionId xid);
extern XLogRecPtr pa_get_last_commit_end(TransactionId xid, bool delete_entry,
bool *skipped_write);
diff --git a/src/test/subscription/t/050_parallel_apply.pl b/src/test/subscription/t/050_parallel_apply.pl
index a6343d57bc5..c90a72682e5 100644
--- a/src/test/subscription/t/050_parallel_apply.pl
+++ b/src/test/subscription/t/050_parallel_apply.pl
@@ -456,4 +456,75 @@ $result =
$node_subscriber->safe_psql('postgres', "SELECT count(1) FROM regress_tab");
is ($result, 5011, 'inserts are replicated to subscriber');
+##################################################
+# Test that mutable user-defined triggers force apply-time waiting so
+# conflicting trigger side effects are serialized.
+##################################################
+
+# Truncate the data for upcoming tests
+$node_publisher->safe_psql('postgres', "TRUNCATE TABLE regress_tab;");
+$node_publisher->wait_for_catchup('regress_sub');
+
+$node_subscriber->safe_psql('postgres', qq[
+ CREATE OR REPLACE FUNCTION regress_trigger_guard_fn()
+ RETURNS trigger
+ LANGUAGE plpgsql
+ AS \$\$
+ BEGIN
+ IF NOT EXISTS (SELECT 1 FROM public.regress_tab WHERE id = 1) THEN
+ UPDATE public.regress_tab SET id = 2 WHERE id = 1;
+ END IF;
+ RETURN NEW;
+ END;
+ \$\$;
+
+ CREATE TRIGGER regress_trigger_guard_tg
+ BEFORE INSERT ON regress_tab
+ FOR EACH ROW
+ EXECUTE FUNCTION regress_trigger_guard_fn();
+]);
+
+# Ensure trigger fires during logical replication apply.
+$node_subscriber->safe_psql('postgres',
+ "ALTER TABLE regress_tab ENABLE REPLICA TRIGGER regress_trigger_guard_tg;"
+);
+
+# Hold the first parallel worker just before commit.
+$node_subscriber->safe_psql('postgres',
+ "SELECT injection_points_attach('parallel-worker-before-commit','wait');"
+);
+
+# TX-1: first insert; worker pauses at before-commit injection point.
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO regress_tab VALUES (1);");
+
+$node_subscriber->wait_for_event('logical replication parallel worker',
+ 'parallel-worker-before-commit');
+
+$offset = -s $node_subscriber->logfile;
+
+# TX-2: second insert. For unsafe trigger tables, this must wait before apply.
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO regress_tab VALUES (2);");
+
+$node_subscriber->wait_for_log(qr/found parallel unsafe change on table [1-9][0-9]* for action [0-9]+/, $offset);
+
+$str = $node_subscriber->wait_for_log(qr/wait for depended xid ([1-9][0-9]+)/, $offset);
+$xid = $str =~ /wait for depended xid ([1-9][0-9]+)/;
+
+ok(1, "mutable trigger relation waits before apply in parallel mode");
+
+# Resume workers.
+$node_subscriber->safe_psql('postgres', qq[
+ SELECT injection_points_detach('parallel-worker-before-commit');
+ SELECT injection_points_wakeup('parallel-worker-before-commit');
+]);
+
+$node_subscriber->wait_for_log(qr/finish waiting for depended xid $xid/, $offset);
+$node_publisher->wait_for_catchup('regress_sub');
+
+$result =
+ $node_subscriber->safe_psql('postgres', "SELECT count(1) FROM regress_tab");
+is ($result, 2, 'inserts are replicated to subscriber');
+
done_testing();
--
2.43.0
[application/octet-stream] v21-0008-Support-dependency-tracking-via-local-unique-ind.patch (39.8K, ../TY4PR01MB177180394D8E9C6D263E117E494FE2@TY4PR01MB17718.jpnprd01.prod.outlook.com/5-v21-0008-Support-dependency-tracking-via-local-unique-ind.patch)
download | inline diff:
From ad1cdc213f4ac7764295d2de1038077ac96ea326 Mon Sep 17 00:00:00 2001
From: Zhijie Hou <[email protected]>
Date: Thu, 9 Jul 2026 15:43:06 +0800
Subject: [PATCH v21 8/9] Support dependency tracking via local unique indexes
This patch tracks dependencies on transactions that modify the same local unique
key. Even if the replica identity keys differ, unique indexes can still cause
conflicts. This is necessary to prevent unexpected errors. For example:
TX-1: DELETE row (1,2) with replica identity key (1,2) and unique key (2)
TX-2: INSERT row (3,2) with replica identity key (3,2) and unique key (2)
If applied in parallel, TX-2's INSERT could be applied before TX-1's DELETE,
leading to a unique index violation error.
We do not track dependencies for INSERT and UPDATE that conflict on a new
unique key value, since such conflicts would cause an error even in serial
mode. Instead, we only track dependencies involving old tuples (from DELETE
or UPDATE) and require INSERT and UPDATE transactions that target the same
unique key to wait for them.
Note that the old tuple of an UPDATE or DELETE may not include the unique key
column if that column is not part of the replica identity columns on the
publisher. In such cases, we only use the unique columns that are part of the
replica identity keys for dependency tracking, which may lead to false
positives. For example, consider a unique index defined as UNIQUE (a, b),
where only b is part of the replica identity keys:
TX-1: DELETE row (1,2) TX-2: INSERT row (3,2)
If applied in parallel, both transactions will be treated as dependent
because they modify the same unique key value (b=2), even though they
actually modify different unique keys. This is acceptable because it is still
better than completely disallowing parallelism for these transactions.
In the worst case, if none of the unique index columns are part of the
replica identity keys, we treat all transactions that modify the same table
as dependent and disallow parallelism for that table.
XXX We could consider requesting the publisher to include unique key columns
in the old tuple of UPDATE or DELETE when they are not part of the replica
identity keys. This would reduce false positives, but would require changes
on the publisher side and increase disk (WAL size) and network data. For now,
we choose not to implement this. An alternative approach is to provide an
option to skip tracking dependencies on unique keys that are not part of the
replica identity keys. This could be useful for users who prefer higher
parallelism and experience few conflicts.
Note that the local unique key could change after dependency checking and
before applying the change. However, to centralize tracking and keep it
simple, we still perform this check only in the leader apply worker. This is
acceptable because in the worst case, the parallel worker will report an
error and restart the transaction using the latest index information.
---
.../replication/logical/applyparallelworker.c | 50 +++
src/backend/replication/logical/relation.c | 179 ++++++++
src/backend/replication/logical/worker.c | 414 +++++++++++++-----
src/backend/storage/lmgr/deadlock.c | 1 -
src/include/replication/logicalrelation.h | 16 +
src/test/subscription/t/050_parallel_apply.pl | 124 ++++++
src/tools/pgindent/typedefs.list | 2 +
7 files changed, 673 insertions(+), 113 deletions(-)
diff --git a/src/backend/replication/logical/applyparallelworker.c b/src/backend/replication/logical/applyparallelworker.c
index 3e436b518bd..1f1a75b5019 100644
--- a/src/backend/replication/logical/applyparallelworker.c
+++ b/src/backend/replication/logical/applyparallelworker.c
@@ -174,6 +174,56 @@
* both are allowed to apply in parallel, TX-2's DELETE could be applied before
* TX-1's INSERT, resulting in a delete_missing conflict.
*
+ * Beyond replica identity keys, we also track dependencies on transactions that
+ * modify the same local unique key. Even if the replica identity keys differ,
+ * unique indexes can still cause conflicts. This is necessary to prevent
+ * unexpected errors. For example:
+ *
+ * TX-1: DELETE row (1,2) with replica identity key (1,2) and unique key (2)
+ * TX-2: INSERT row (3,2) with replica identity key (3,2) and unique key (2)
+ *
+ * If applied in parallel, TX-2's INSERT could be applied before TX-1's DELETE,
+ * leading to a unique index violation error.
+ *
+ * We do not track dependencies for INSERT and UPDATE that conflict on a new
+ * unique key value, since such conflicts would cause an error even in serial
+ * mode. Instead, we only track dependencies involving old tuples (from DELETE
+ * or UPDATE) and require INSERT and UPDATE transactions that target the same
+ * unique key to wait for them.
+ *
+ * Note that the old tuple of an UPDATE or DELETE may not include the unique key
+ * column if that column is not part of the replica identity columns on the
+ * publisher. In such cases, we only use the unique columns that are part of the
+ * replica identity keys for dependency tracking, which may lead to false
+ * positives. For example, consider a unique index defined as UNIQUE (a, b),
+ * where only b is part of the replica identity keys:
+ *
+ * TX-1: DELETE row (1,2) TX-2: INSERT row (3,2)
+ *
+ * If applied in parallel, both transactions will be treated as dependent
+ * because they modify the same unique key value (b=2), even though they
+ * actually modify different unique keys. This is acceptable because it is still
+ * better than completely disallowing parallelism for these transactions.
+ *
+ * In the worst case, if none of the unique index columns are part of the
+ * replica identity keys, we treat all transactions that modify the same table
+ * as dependent and disallow parallelism for that table.
+ *
+ * XXX We could consider requesting the publisher to include unique key columns
+ * in the old tuple of UPDATE or DELETE when they are not part of the replica
+ * identity keys. This would reduce false positives, but would require changes
+ * on the publisher side and increase disk (WAL size) and network data. For now,
+ * we choose not to implement this. An alternative approach is to provide an
+ * option to skip tracking dependencies on unique keys that are not part of the
+ * replica identity keys. This could be useful for users who prefer higher
+ * parallelism and experience few conflicts.
+ *
+ * Note that the local unique key could change after dependency checking and
+ * before applying the change. However, to centralize tracking and keep it
+ * simple, we still perform this check only in the leader apply worker. This is
+ * acceptable because in the worst case, the parallel worker will report an
+ * error and restart the transaction using the latest index information.
+ *
* Commit order
* ------------
* We preserve publisher commit order for all transactions for two reasons:
diff --git a/src/backend/replication/logical/relation.c b/src/backend/replication/logical/relation.c
index 6d3e1827ae2..1572990ee18 100644
--- a/src/backend/replication/logical/relation.c
+++ b/src/backend/replication/logical/relation.c
@@ -88,6 +88,7 @@ logicalrep_relmap_invalidate_cb(Datum arg, Oid reloid)
{
entry->localrelvalid = false;
entry->parallel_safety_valid = false;
+ entry->local_unique_indexes_valid = false;
hash_seq_term(&status);
break;
}
@@ -104,6 +105,7 @@ logicalrep_relmap_invalidate_cb(Datum arg, Oid reloid)
{
entry->localrelvalid = false;
entry->parallel_safety_valid = false;
+ entry->local_unique_indexes_valid = false;
}
}
}
@@ -135,6 +137,21 @@ logicalrep_relmap_init(void)
(Datum) 0);
}
+/*
+ * Release local index list
+ */
+static void
+free_local_unique_indexes(LogicalRepRelMapEntry *entry)
+{
+ Assert(am_leader_apply_worker());
+
+ foreach_ptr(LogicalRepSubUniqueIndex, idxinfo, entry->local_unique_indexes)
+ bms_free(idxinfo->indexkeys);
+
+ list_free_deep(entry->local_unique_indexes);
+ entry->local_unique_indexes = NIL;
+}
+
/*
* Free the entry of a relation map cache.
*/
@@ -162,6 +179,9 @@ logicalrep_relmap_free_entry(LogicalRepRelMapEntry *entry)
if (entry->attrmap)
free_attrmap(entry->attrmap);
+
+ if (entry->local_unique_indexes != NIL)
+ free_local_unique_indexes(entry);
}
/*
@@ -218,6 +238,13 @@ logicalrep_relmap_update(LogicalRepRelation *remoterel)
(remoterel->relkind == 0) ? RELKIND_RELATION : remoterel->relkind;
entry->remoterel.attkeys = bms_copy(remoterel->attkeys);
+
+ /*
+ * Rebuild the key info using the latest replica identity, which may have
+ * changed.
+ */
+ entry->local_unique_indexes_valid = false;
+
MemoryContextSwitchTo(oldctx);
}
@@ -404,6 +431,7 @@ logicalrep_rel_open(LogicalRepRelId remoteid, LOCKMODE lockmode)
/* Table was renamed or dropped. */
entry->localrelvalid = false;
entry->parallel_safety_valid = false;
+ entry->local_unique_indexes_valid = false;
}
else if (!entry->localrelvalid)
{
@@ -614,6 +642,157 @@ logicalrep_rel_check_parallel_safety(LogicalRepRelMapEntry *entry)
entry->parallel_safety_valid = true;
}
+/*
+ * Collect all local unique indexes that can be used for dependency tracking
+ *
+ * This function collects all types of unique indexes, including those with
+ * index expressions and partial indexes. However, to avoid the overhead and
+ * complexity of executing expressions, we do not evaluate them during
+ * dependency tracking.
+ *
+ * For indexes with expressions, only the non-expression columns are recorded in
+ * the bitmap. The dependency tracking function will use only these columns,
+ * which may lead to false dependency detection. For example, consider a unique
+ * index defined as UNIQUE (a, func(b)), where b is an expression column. Rows
+ * (1, 2) and (1, 3) will be treated as dependent even though they are not. This
+ * is acceptable, as it is still better than disabling parallelism for all
+ * relations that have expression indexes.
+ *
+ * Similarly, partial indexes may also cause false dependencies due to predicate
+ * expressions. For the same reason, we consider this acceptable as well.
+ *
+ * To avoid redundant dependency tracking, indexes whose key columns are the
+ * same as, or a superset of, the replica identity key columns are skipped,
+ * since tracking the replica identity keys already covers their scope.
+ *
+ * Columns not in the replica identity key are excluded from the unique column
+ * set. Since the old tuple of an UPDATE or DELETE contains only replica
+ * identity key columns, any other columns would be missing and thus unavailable
+ * for dependency tracking.
+ */
+void
+logicalrep_build_dependent_unique_indexes(LogicalRepRelMapEntry *entry)
+{
+ List *idxlist;
+
+ if (entry->local_unique_indexes_valid)
+ return;
+
+ free_local_unique_indexes(entry);
+
+ /*
+ * XXX For partitioned tables, we must collect unique indexes from leaf
+ * partitions, which are the actual replication targets. This is because
+ * leaf partitions can have unique indexes that are not present on the
+ * partitioned table, and those indexes can be used for dependency tracking.
+ * However, collecting unique indexes from leaf partitions requires building
+ * the tuple, and executing partition pruning expressions, which could be
+ * expensive for each change on a partitioned table. For now, we skip
+ * collecting local unique indexes for partitioned tables and create a dummy
+ * entry, ensuring that changes on partitioned tables are not applied in
+ * parallel.
+ */
+ if (entry->localrel->rd_rel->relkind == RELKIND_PARTITIONED_TABLE)
+ {
+ MemoryContext oldctx;
+ LogicalRepSubUniqueIndex *indexinfo;
+
+ oldctx = MemoryContextSwitchTo(LogicalRepRelMapContext);
+ indexinfo = palloc(sizeof(LogicalRepSubUniqueIndex));
+ indexinfo->indexoid = InvalidOid;
+ indexinfo->indexkeys = NULL;
+ indexinfo->nulls_distinct = false;
+ entry->local_unique_indexes = lappend(entry->local_unique_indexes,
+ indexinfo);
+ MemoryContextSwitchTo(oldctx);
+
+ entry->local_unique_indexes_valid = true;
+
+ return;
+ }
+
+ idxlist = RelationGetIndexList(entry->localrel);
+
+ /* Iterate indexes to list all usable indexes */
+ foreach_oid(idxoid, idxlist)
+ {
+ Relation idxrel;
+ int indnkeys;
+ AttrMap *attrmap;
+ MemoryContext oldctx;
+ LogicalRepSubUniqueIndex *indexinfo;
+ Bitmapset *indexkeys = NULL;
+ bool nulls_distinct;
+
+ idxrel = index_open(idxoid, AccessShareLock);
+
+ /* Only unique indexes are considered */
+ if (!idxrel->rd_index->indisunique)
+ {
+ index_close(idxrel, AccessShareLock);
+ continue;
+ }
+
+ indnkeys = idxrel->rd_index->indnkeyatts;
+ nulls_distinct = !idxrel->rd_index->indnullsnotdistinct;
+ attrmap = entry->attrmap;
+
+ Assert(indnkeys);
+
+ /* Seek each attributes and add to a Bitmap */
+ for (int i = 0; i < indnkeys; i++)
+ {
+ AttrNumber localcol = idxrel->rd_index->indkey.values[i];
+ AttrNumber remotecol;
+
+ /* Skip expression */
+ if (!AttributeNumberIsValid(localcol))
+ continue;
+
+ remotecol = attrmap->attnums[AttrNumberGetAttrOffset(localcol)];
+
+ /* Skip if the column does not exist on publisher node */
+ if (remotecol < 0)
+ continue;
+
+ /* Skip columns that are not part of the replica identity key */
+ if (!bms_is_member(remotecol, entry->remoterel.attkeys))
+ continue;
+
+ /* Checks are passed, remember the attribute */
+ indexkeys = bms_add_member(indexkeys, remotecol);
+ }
+
+ index_close(idxrel, AccessShareLock);
+
+ /*
+ * Skip indexes whose key columns are a superset of the replica identity
+ * key.
+ */
+ if (bms_equal(entry->remoterel.attkeys, indexkeys) ||
+ bms_is_subset(entry->remoterel.attkeys, indexkeys))
+ {
+ bms_free(indexkeys);
+ continue;
+ }
+
+ oldctx = MemoryContextSwitchTo(LogicalRepRelMapContext);
+ indexinfo = palloc(sizeof(LogicalRepSubUniqueIndex));
+ indexinfo->indexoid = idxoid;
+ indexinfo->indexkeys = bms_copy(indexkeys);
+ indexinfo->nulls_distinct = nulls_distinct;
+ entry->local_unique_indexes = lappend(entry->local_unique_indexes,
+ indexinfo);
+ MemoryContextSwitchTo(oldctx);
+
+ bms_free(indexkeys);
+ }
+
+ list_free(idxlist);
+
+ entry->local_unique_indexes_valid = true;
+}
+
/*
* Partition cache: look up partition LogicalRepRelMapEntry's
*
diff --git a/src/backend/replication/logical/worker.c b/src/backend/replication/logical/worker.c
index 18d71ceabe6..3864c95f930 100644
--- a/src/backend/replication/logical/worker.c
+++ b/src/backend/replication/logical/worker.c
@@ -578,10 +578,20 @@ typedef struct ApplySubXactData
static ApplySubXactData subxact_data = {0, 0, InvalidTransactionId, NULL};
+/*
+ * Type of key used for dependency tracking.
+ */
+typedef enum LogicalRepKeyKind
+{
+ LOGICALREP_KEY_REPLICA_IDENTITY,
+ LOGICALREP_KEY_LOCAL_UNIQUE
+} LogicalRepKeyKind;
+
/* Hash table key for replica_identity_table */
typedef struct ReplicaIdentityKey
{
Oid relid;
+ LogicalRepKeyKind kind;
LogicalRepTupleData *data;
} ReplicaIdentityKey;
@@ -761,7 +771,8 @@ static bool
hash_replica_identity_compare(ReplicaIdentityKey *a, ReplicaIdentityKey *b)
{
if (a->relid != b->relid ||
- a->data->ncols != b->data->ncols)
+ a->data->ncols != b->data->ncols ||
+ a->kind != b->kind)
return false;
for (int i = 0; i < a->data->ncols; i++)
@@ -792,8 +803,12 @@ free_replica_identity_key(ReplicaIdentityKey *key)
{
Assert(key);
- pfree(key->data->colvalues);
- pfree(key->data->colstatus);
+ if (key->data->colvalues)
+ pfree(key->data->colvalues);
+
+ if (key->data->colstatus)
+ pfree(key->data->colstatus);
+
pfree(key->data);
pfree(key);
}
@@ -935,6 +950,79 @@ append_xid_dependency(TransactionId xid, List **depends_on_xids)
*depends_on_xids = lappend_xid(*depends_on_xids, xid);
}
+/*
+ * Common function for checking dependency by using the key. Used by both
+ * check_and_record_ri_dependency and check_and_record_local_key_dependency.
+ *
+ * Check whether the given key has an active dependency. If new_depended_xid is
+ * valid, also records a new dependency for that transaction.
+ *
+ * Return the existing transaction ID if an active dependency exists for the
+ * key; otherwise returns InvalidTransactionId.
+ */
+static TransactionId
+check_and_record_key_dependency(ReplicaIdentityKey *key,
+ TransactionId new_depended_xid)
+{
+ TransactionId existing_xid = InvalidTransactionId;
+ ReplicaIdentityEntry *rientry;
+ bool found = false;
+
+ /*
+ * The new xid could be invalid if the transaction will be applied by the
+ * leader itself which means all the changes will be committed before
+ * processing next transaction. In this case, we only need to check for
+ * dependencies on preceding transactions, there is no need to record a new
+ * dependency for subsequent transactions to wait on.
+ */
+ if (!TransactionIdIsValid(new_depended_xid))
+ {
+ rientry = replica_identity_lookup(replica_identity_table, key);
+
+ if (rientry && has_active_key_dependency(rientry, true))
+ {
+ elog(DEBUG1,
+ key->kind == LOGICALREP_KEY_REPLICA_IDENTITY ?
+ "found conflicting replica identity change on table %u from %u" :
+ "found conflicting local unique key change on table %u from %u",
+ key->relid, rientry->remote_xid);
+
+ existing_xid = rientry->remote_xid;
+ }
+
+ free_replica_identity_key(key);
+
+ return existing_xid;
+ }
+
+ /* Record a new dependency for subsequent transactions to wait on */
+ rientry = replica_identity_insert(replica_identity_table, key,
+ &found);
+
+ /*
+ * Release the key built to search the entry, if the entry already exists.
+ */
+ if (found)
+ {
+ if (has_active_key_dependency(rientry, false))
+ {
+ elog(DEBUG1,
+ key->kind == LOGICALREP_KEY_REPLICA_IDENTITY ?
+ "found conflicting replica identity change on table %u from %u" :
+ "found conflicting local unique key change on table %u from %u",
+ key->relid, rientry->remote_xid);
+
+ existing_xid = rientry->remote_xid;
+ }
+
+ free_replica_identity_key(key);
+ }
+
+ rientry->remote_xid = new_depended_xid;
+
+ return existing_xid;
+}
+
/*
* Check if any of the key columns have NULL values.
*/
@@ -952,81 +1040,34 @@ has_null_key_values(LogicalRepTupleData *data, Bitmapset *indexkeys)
}
/*
- * Check for dependencies on preceding transactions that modify the same key as
- * the given tuple. Returns the dependent transactions in 'depends_on_xids'.
- *
- * Additionally, if new_depended_xid is valid, record the current change and the
- * transaction as a new dependency for the replica identity key modification,
- * allowing subsequent transactions that modify the same key to be dependent on
- * it.
+ * Build a hash key for replica_identity_table using the given relation and
+ * tuple data, restricted to the specified key columns.
*/
-static void
-check_and_record_ri_dependency(Oid relid, LogicalRepTupleData *original_data,
- TransactionId new_depended_xid,
- List **depends_on_xids)
+static ReplicaIdentityKey *
+build_replica_identity_key(Oid relid, LogicalRepTupleData *original_data,
+ Bitmapset *keycols)
{
- LogicalRepRelMapEntry *relentry;
- LogicalRepTupleData *ridata;
- ReplicaIdentityKey *rikey;
- ReplicaIdentityEntry *rientry;
+ LogicalRepTupleData *keydata;
+ ReplicaIdentityKey *key;
MemoryContext oldctx;
- int n_ri;
- bool found = false;
-
- Assert(depends_on_xids);
+ int nkeycols = bms_num_members(keycols);
- /* Search for existing entry */
- relentry = logicalrep_get_relentry(relid);
+ oldctx = MemoryContextSwitchTo(ApplyContext);
- Assert(relentry);
+ /* Allocate space for replica identity values */
+ keydata = palloc0_object(LogicalRepTupleData);
- /*
- * First check whether any previous transaction (other than the current one)
- * has affected the whole table e.g., truncate or schema change from
- * publisher.
- */
- if (has_active_rel_dependency(relentry) &&
- !TransactionIdEquals(relentry->last_depended_xid, new_depended_xid))
+ if (nkeycols)
{
- elog(DEBUG1, "found table-wide change affecting %u from %u",
- relid, relentry->last_depended_xid);
-
- append_xid_dependency(relentry->last_depended_xid, depends_on_xids);
+ keydata->colvalues = palloc0_array(StringInfoData, nkeycols);
+ keydata->colstatus = palloc0_array(char, nkeycols);
}
- n_ri = bms_num_members(relentry->remoterel.attkeys);
-
- /*
- * Return if there are no replica identity columns, indicating that the
- * remote relation has neither a replica identity key nor is marked as
- * replica identity full.
- */
- if (!n_ri)
- return;
-
- /*
- * For replica identity FULL, the tuple contains all columns including
- * NULLs, so we always check for existing dependencies.
- *
- * For other cases, NULL in the new tuple means the replica identity key
- * hasn't changed, so no new dependency needs to be recorded. The dependency
- * should have been recorded when processing the old tuple.
- */
- if (relentry->remoterel.replident != REPLICA_IDENTITY_FULL &&
- has_null_key_values(original_data, relentry->remoterel.attkeys))
- return;
-
- oldctx = MemoryContextSwitchTo(ApplyContext);
+ keydata->ncols = nkeycols;
- /* Allocate space for replica identity values */
- ridata = palloc0_object(LogicalRepTupleData);
- ridata->colvalues = palloc0_array(StringInfoData, n_ri);
- ridata->colstatus = palloc0_array(char, n_ri);
- ridata->ncols = n_ri;
-
- for (int i_original = 0, i_ri = 0; i_original < original_data->ncols; i_original++)
+ for (int i = 0, i_key = 0; i < original_data->ncols; i++)
{
- if (!bms_is_member(i_original, relentry->remoterel.attkeys))
+ if (!bms_is_member(i, keycols))
continue;
/*
@@ -1039,77 +1080,206 @@ check_and_record_ri_dependency(Oid relid, LogicalRepTupleData *original_data,
* see the complete replica identity key value in original_data and
* correctly check the dependency.
*/
- Assert(original_data->colstatus[i_original] != LOGICALREP_COLUMN_UNCHANGED ||
- original_data->colvalues[i_original].len > 0);
+ Assert(original_data->colstatus[i] != LOGICALREP_COLUMN_UNCHANGED ||
+ original_data->colvalues[i].len > 0);
- if (original_data->colstatus[i_original] != LOGICALREP_COLUMN_NULL)
+ if (original_data->colstatus[i] != LOGICALREP_COLUMN_NULL)
{
- StringInfo original_colvalue = &original_data->colvalues[i_original];
+ StringInfo original_colvalue = &original_data->colvalues[i];
- initStringInfoExt(&ridata->colvalues[i_ri], original_colvalue->len + 1);
- appendStringInfoString(&ridata->colvalues[i_ri], original_colvalue->data);
+ initStringInfoExt(&keydata->colvalues[i_key], original_colvalue->len + 1);
+ appendStringInfoString(&keydata->colvalues[i_key], original_colvalue->data);
}
- ridata->colstatus[i_ri] = original_data->colstatus[i_original];
- i_ri++;
+ keydata->colstatus[i_key] = original_data->colstatus[i];
+ i_key++;
}
- rikey = palloc0_object(ReplicaIdentityKey);
- rikey->relid = relid;
- rikey->data = ridata;
+ key = palloc0_object(ReplicaIdentityKey);
+ key->relid = relid;
+ key->data = keydata;
MemoryContextSwitchTo(oldctx);
+ return key;
+}
+
+/*
+ * Build dependency key information for the given relation entry if not already
+ * collected.
+ *
+ * See logicalrep_build_dependent_unique_indexes() for details.
+ */
+static void
+build_local_dependent_key_info(LogicalRepRelMapEntry *relentry)
+{
+ bool needs_start;
+
+ if (relentry->local_unique_indexes_valid)
+ return;
+
/*
- * The new xid could be invalid if the transaction will be applied by the
- * leader itself which means all the changes will be committed before
- * processing next transaction. In this case, we only need to check for
- * dependencies on preceding transactions, there is no need to record a new
- * dependency for subsequent transactions to wait on.
+ * Gather information for local indexes if not yet. We require to be in a
+ * transaction state to collect indexes info from system catalogs.
*/
- if (!TransactionIdIsValid(new_depended_xid))
- {
- rientry = replica_identity_lookup(replica_identity_table, rikey);
- free_replica_identity_key(rikey);
+ needs_start = !IsTransactionState();
- if (rientry && has_active_key_dependency(rientry, true))
- {
- elog(DEBUG1, "found conflicting replica identity change on table %u from %u",
- relid, rientry->remote_xid);
+ if (needs_start)
+ StartTransactionCommand();
- append_xid_dependency(rientry->remote_xid, depends_on_xids);
- }
+ relentry = logicalrep_rel_open(relentry->remoterel.remoteid, AccessShareLock);
- return;
- }
+ logicalrep_build_dependent_unique_indexes(relentry);
- /* Record a new dependency for subsequent transactions to wait on */
- rientry = replica_identity_insert(replica_identity_table, rikey,
- &found);
+ logicalrep_rel_close(relentry, AccessShareLock);
- /*
- * Release the key built to search the entry, if the entry already exists.
- */
- if (found)
+ if (needs_start)
+ CommitTransactionCommand();
+}
+
+/*
+ * Mostly same as check_and_record_ri_dependency() but for local unique indexes.
+ *
+ * See the comments in applyparallelworker.c for details on why tracking these
+ * dependencies is necessary.
+ */
+static void
+check_and_record_local_key_dependency(Oid relid,
+ LogicalRepTupleData *original_data,
+ bool old_tuple,
+ TransactionId new_depended_xid,
+ List **depends_on_xids)
+{
+ LogicalRepRelMapEntry *relentry;
+ ReplicaIdentityKey *rikey;
+
+ Assert(depends_on_xids);
+
+ /* Search for existing entry */
+ relentry = logicalrep_get_relentry(relid);
+
+ Assert(relentry);
+
+ build_local_dependent_key_info(relentry);
+
+ foreach_ptr(LogicalRepSubUniqueIndex, idxinfo, relentry->local_unique_indexes)
{
- free_replica_identity_key(rikey);
+ /*
+ * NULL values in the new tuple represent true NULLs in a unique index.
+ * If NULLs are treated as distinct (nulls_distinct = true), they never
+ * cause conflicts. Therefore, we can skip dependency checking if any
+ * key column is NULL in this case.
+ *
+ * However, for old tuples in UPDATE or DELETE operations, a NULL key
+ * simply indicate the column lies outside the replica identity key
+ * rather than a true NULL. In such cases, the remote old tuple could
+ * still conflict with a local tuple, so we must not skip the check.
+ */
+ if (!old_tuple && idxinfo->nulls_distinct &&
+ has_null_key_values(original_data, idxinfo->indexkeys))
+ continue;
/*
- * Append the dependency to the list if the current transaction was not
- * the lastest one to modify the key.
+ * Old tuples of unique keys do not conflict with any preceding
+ * transaction (see the comments in applyparallelworker.c for details on
+ * conflicting cases). When we don't need to record a new dependency, we
+ * can skip processing this index entirely.
*/
- if (has_active_key_dependency(rientry, false) &&
- !TransactionIdEquals(rientry->remote_xid, new_depended_xid))
+ if (old_tuple && !TransactionIdIsValid(new_depended_xid))
+ continue;
+
+ rikey = build_replica_identity_key(relid, original_data, idxinfo->indexkeys);
+ rikey->kind = LOGICALREP_KEY_LOCAL_UNIQUE;
+
+ /*
+ * For old tuples, record a dependency for subsequent transactions to
+ * wait on; no preceding transactions are added to the list.
+ *
+ * For new tuples in INSERT or UPDATE, check for existing key
+ * dependencies and add any dependent transactions to the list.
+ */
+ if (old_tuple)
+ {
+ (void) check_and_record_key_dependency(rikey, new_depended_xid);
+ }
+ else
{
- elog(DEBUG1, "found conflicting replica identity change on table %u from %u",
- relid, rientry->remote_xid);
+ TransactionId xid;
+
+ xid = check_and_record_key_dependency(rikey, InvalidTransactionId);
- append_xid_dependency(rientry->remote_xid, depends_on_xids);
+ if (TransactionIdIsValid(xid) &&
+ !TransactionIdEquals(xid, new_depended_xid))
+ append_xid_dependency(xid, depends_on_xids);
}
}
+}
- /* Update the new depended xid into the entry */
- rientry->remote_xid = new_depended_xid;
+/*
+ * Check for dependencies on preceding transactions that modify the same key.
+ * Returns the dependent transactions in 'depends_on_xids'.
+ *
+ * Additionally, if new_depended_xid is valid, record it as a dependency for the
+ * replica identity key modification, allowing subsequent transactions that
+ * modify the same key to be dependent on it.
+ */
+static void
+check_and_record_ri_dependency(Oid relid, LogicalRepTupleData *original_data,
+ TransactionId new_depended_xid,
+ List **depends_on_xids)
+{
+ LogicalRepRelMapEntry *relentry;
+ ReplicaIdentityKey *rikey;
+ TransactionId xid;
+
+ Assert(depends_on_xids);
+
+ /* Search for existing entry */
+ relentry = logicalrep_get_relentry(relid);
+
+ Assert(relentry);
+
+ /*
+ * First check whether any previous transaction (other than the current one)
+ * has affected the whole table e.g., truncate or schema change from
+ * publisher.
+ */
+ if (has_active_rel_dependency(relentry) &&
+ !TransactionIdEquals(relentry->last_depended_xid, new_depended_xid))
+ append_xid_dependency(relentry->last_depended_xid, depends_on_xids);
+
+ /*
+ * Return if there are no replica identity columns, indicating that the
+ * remote relation has neither a replica identity key nor is marked as
+ * replica identity full.
+ */
+ if (!bms_num_members(relentry->remoterel.attkeys))
+ return;
+
+ /*
+ * For replica identity FULL, the tuple contains all columns including
+ * NULLs, so we always check for existing dependencies.
+ *
+ * For other cases, NULL in the new tuple means the replica identity key
+ * hasn't changed, so no new dependency needs to be recorded. The dependency
+ * should have been recorded when processing the old tuple.
+ */
+ if (REPLICA_IDENTITY_FULL != relentry->remoterel.replident &&
+ has_null_key_values(original_data, relentry->remoterel.attkeys))
+ return;
+
+ rikey = build_replica_identity_key(relid, original_data, relentry->remoterel.attkeys);
+ rikey->kind = LOGICALREP_KEY_REPLICA_IDENTITY;
+
+ xid = check_and_record_key_dependency(rikey, new_depended_xid);
+
+ /*
+ * Append the dependency to the list if the current transaction was not the
+ * lastest one to modify the key.
+ */
+ if (TransactionIdIsValid(xid) &&
+ !TransactionIdEquals(xid, new_depended_xid))
+ append_xid_dependency(xid, depends_on_xids);
}
/*
@@ -1308,6 +1478,9 @@ handle_dependency_on_change(LogicalRepMsgType action, StringInfo s,
relid = logicalrep_read_insert(&change, &newtup);
check_and_record_ri_dependency(relid, &newtup, new_depended_xid,
&depends_on_xids);
+ check_and_record_local_key_dependency(relid, &newtup, false,
+ new_depended_xid,
+ &depends_on_xids);
break;
case LOGICAL_REP_MSG_UPDATE:
@@ -1337,12 +1510,29 @@ handle_dependency_on_change(LogicalRepMsgType action, StringInfo s,
check_and_record_ri_dependency(relid, &newtup, new_depended_xid,
&depends_on_xids);
+
+ /*
+ * Check the new tuple first to detect dependencies on preceding
+ * transactions that modified the same key. If we processed the old
+ * tuple first, it might update the same hash entry with the current
+ * transaction ID, causing the new tuple check to miss any preceding
+ * transaction.
+ */
+ check_and_record_local_key_dependency(relid, &newtup, false,
+ new_depended_xid,
+ &depends_on_xids);
+ check_and_record_local_key_dependency(relid, &oldtup, true,
+ new_depended_xid,
+ &depends_on_xids);
break;
case LOGICAL_REP_MSG_DELETE:
relid = logicalrep_read_delete(&change, &oldtup);
check_and_record_ri_dependency(relid, &oldtup, new_depended_xid,
&depends_on_xids);
+ check_and_record_local_key_dependency(relid, &oldtup, true,
+ new_depended_xid,
+ &depends_on_xids);
break;
case LOGICAL_REP_MSG_TRUNCATE:
diff --git a/src/backend/storage/lmgr/deadlock.c b/src/backend/storage/lmgr/deadlock.c
index b8962d875b6..31bbfcf1971 100644
--- a/src/backend/storage/lmgr/deadlock.c
+++ b/src/backend/storage/lmgr/deadlock.c
@@ -33,7 +33,6 @@
#include "storage/procnumber.h"
#include "utils/memutils.h"
-
/*
* One edge in the waits-for graph.
*
diff --git a/src/include/replication/logicalrelation.h b/src/include/replication/logicalrelation.h
index 439243be742..ec4c9ece571 100644
--- a/src/include/replication/logicalrelation.h
+++ b/src/include/replication/logicalrelation.h
@@ -56,6 +56,10 @@ typedef struct LogicalRepRelMapEntry
*/
TransactionId last_depended_xid;
+ /* Local unique indexes. Used for dependency tracking */
+ List *local_unique_indexes;
+ bool local_unique_indexes_valid;
+
/*
* Per-operation safety cache for parallel apply. If
* parallel_global_unsafe[action] is true, that action cannot be applied in
@@ -74,6 +78,17 @@ typedef struct LogicalRepRelMapEntry
bool parallel_global_unsafe[LRPA_ACTION_COUNT];
} LogicalRepRelMapEntry;
+/*
+ * Subscriber side unique index information. This is used to track dependencies
+ * between transactions that modify the same unique key value.
+ */
+typedef struct LogicalRepSubUniqueIndex
+{
+ Oid indexoid; /* OID of the local key */
+ Bitmapset *indexkeys; /* Bitmap of key columns *on remote* */
+ bool nulls_distinct; /* Whether NULLs are considered distinct */
+} LogicalRepSubUniqueIndex;
+
extern void logicalrep_relmap_update(LogicalRepRelation *remoterel);
extern void logicalrep_partmap_reset_relmap(LogicalRepRelation *remoterel);
@@ -84,6 +99,7 @@ extern LogicalRepRelMapEntry *logicalrep_partition_open(LogicalRepRelMapEntry *r
extern void logicalrep_rel_close(LogicalRepRelMapEntry *rel,
LOCKMODE lockmode);
extern void logicalrep_rel_check_parallel_safety(LogicalRepRelMapEntry *entry);
+extern void logicalrep_build_dependent_unique_indexes(LogicalRepRelMapEntry *entry);
extern bool IsIndexUsableForReplicaIdentityFull(Relation idxrel, AttrMap *attrmap);
extern Oid GetRelationIdentityOrPK(Relation rel);
extern int logicalrep_get_num_rels(void);
diff --git a/src/test/subscription/t/050_parallel_apply.pl b/src/test/subscription/t/050_parallel_apply.pl
index c90a72682e5..9e981bab981 100644
--- a/src/test/subscription/t/050_parallel_apply.pl
+++ b/src/test/subscription/t/050_parallel_apply.pl
@@ -527,4 +527,128 @@ $result =
$node_subscriber->safe_psql('postgres', "SELECT count(1) FROM regress_tab");
is ($result, 2, 'inserts are replicated to subscriber');
+##################################################
+# Test that the dependency tracking works correctly for local unique indexes on
+# subscriber during parallel apply.
+##################################################
+
+# Truncate the data for upcoming tests
+$node_publisher->safe_psql('postgres', "TRUNCATE TABLE regress_tab;");
+$node_publisher->wait_for_catchup('regress_sub');
+
+# Define an unique index on subscriber
+$node_subscriber->safe_psql('postgres',
+ "CREATE UNIQUE INDEX local_unique_idx ON regress_tab (value);");
+
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO regress_tab VALUES (1, 'would conflict');");
+
+$node_publisher->wait_for_catchup('regress_sub');
+
+$result =
+ $node_subscriber->safe_psql('postgres', "SELECT count(1) FROM regress_tab");
+is ($result, 1, 'the insert is replicated to subscriber');
+
+# Attach an injection_point. Parallel workers would wait before the commit
+$node_subscriber->safe_psql('postgres',
+ "SELECT injection_points_attach('parallel-worker-before-commit','wait');"
+);
+
+# Delete the tuple on publisher.
+$node_publisher->safe_psql('postgres',
+ "DELETE FROM regress_tab WHERE id = 1;");
+
+# Wait until the parallel worker enters the injection point.
+$node_subscriber->wait_for_event('logical replication parallel worker',
+ 'parallel-worker-before-commit');
+
+$offset = -s $node_subscriber->logfile;
+
+# Insert tuples. This should conflict with the DELETE transaction, as both
+# transactions modify the same key. The parallel worker will wait for the
+# preceding transaction to finish.
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO regress_tab VALUES (2, 'would conflict');");
+
+# Verify the dependency is detected for the insert
+$str = $node_subscriber->wait_for_log(qr/found conflicting local unique key change on table [1-9][0-9]+ from ([1-9][0-9]+)/, $offset);
+$xid = $str =~ /found conflicting local unique key change on table [1-9][0-9]+ from ([1-9][0-9]+)/;
+
+# Verify the parallel worker waits for the same transaction
+$node_subscriber->wait_for_log(qr/wait for depended xid $xid/, $offset);
+
+ok(1, "local unique key dependency detected for parallel apply");
+
+# Wakeup the parallel worker
+$node_subscriber->safe_psql('postgres', qq[
+ SELECT injection_points_detach('parallel-worker-before-commit');
+ SELECT injection_points_wakeup('parallel-worker-before-commit');
+]);
+
+# Verify the streamed transaction can be applied
+$node_subscriber->wait_for_log(qr/finish waiting for depended xid $xid/, $offset);
+
+$node_publisher->wait_for_catchup('regress_sub');
+
+$result =
+ $node_subscriber->safe_psql('postgres', "SELECT count(1) FROM regress_tab");
+is ($result, 1, 'inserts are replicated to subscriber');
+
+# Test that the dependency tracking works correctly for local unique indexes on
+# subscriber during parallel apply when the unique index has expression.
+$node_subscriber->safe_psql('postgres', "DROP INDEX local_unique_idx;");
+$node_subscriber->safe_psql('postgres',
+ "CREATE UNIQUE INDEX local_unique_idx_expr ON regress_tab ((LOWER(value)));");
+
+# Attach an injection_point. Parallel workers would wait before the commit
+$node_subscriber->safe_psql('postgres',
+ "SELECT injection_points_attach('parallel-worker-before-commit','wait');"
+);
+
+# Insert a tuple on publisher. Parallel worker would wait at the injection
+# point
+$node_publisher->safe_psql('postgres',
+ "DELETE FROM regress_tab WHERE id = 2;");
+
+# Wait until the parallel worker enters the injection point.
+$node_subscriber->wait_for_event('logical replication parallel worker',
+ 'parallel-worker-before-commit');
+
+$offset = -s $node_subscriber->logfile;
+
+# Insert tuples. This should conflict with the DELETE transaction, as both
+# transactions modify the same key value. The parallel worker will wait for the
+# preceding transaction to finish.
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO regress_tab VALUES (3, 'WOULD CONFLICT');");
+
+# Verify the dependency is detected for the insert
+$str = $node_subscriber->wait_for_log(qr/found conflicting local unique key change on table [1-9][0-9]+ from ([1-9][0-9]+)/, $offset);
+$xid = $str =~ /found conflicting local unique key change on table [1-9][0-9]+ from ([1-9][0-9]+)/;
+
+# Verify the parallel worker waits for the same transaction
+$node_subscriber->wait_for_log(qr/wait for depended xid $xid/, $offset);
+
+ok(1, "local unique key dependency from index expression detected for parallel apply");
+
+# Wakeup the parallel worker
+$node_subscriber->safe_psql('postgres', qq[
+ SELECT injection_points_detach('parallel-worker-before-commit');
+ SELECT injection_points_wakeup('parallel-worker-before-commit');
+]);
+
+# Verify the streamed transaction can be applied
+$node_subscriber->wait_for_log(qr/finish waiting for depended xid $xid/, $offset);
+
+$node_publisher->wait_for_catchup('regress_sub');
+
+$result =
+ $node_subscriber->safe_psql('postgres', "SELECT count(1) FROM regress_tab");
+is ($result, 1, 'inserts are replicated to subscriber');
+
+# Cleanup
+$node_subscriber->safe_psql('postgres', "DROP INDEX local_unique_idx_expr;");
+$node_publisher->safe_psql('postgres', "TRUNCATE TABLE regress_tab;");
+$node_publisher->wait_for_catchup('regress_sub');
+
done_testing();
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index c9b24dbfa59..987489f1ef1 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1704,6 +1704,7 @@ LogicalRepBeginData
LogicalRepCommitData
LogicalRepCommitPreparedTxnData
LogicalRepCtxStruct
+LogicalRepKeyKind
LogicalRepMsgType
LogicalRepPartMapEntry
LogicalRepPreparedTxnData
@@ -1713,6 +1714,7 @@ LogicalRepRelation
LogicalRepRollbackPreparedTxnData
LogicalRepSequenceInfo
LogicalRepStreamAbortData
+LogicalRepSubscriberIdx
LogicalRepTupleData
LogicalRepTyp
LogicalRepWorker
--
2.43.0
[application/octet-stream] v21-0006-Track-dependencies-for-streamed-transactions.patch (10.4K, ../TY4PR01MB177180394D8E9C6D263E117E494FE2@TY4PR01MB17718.jpnprd01.prod.outlook.com/6-v21-0006-Track-dependencies-for-streamed-transactions.patch)
download | inline diff:
From cc527f4bf5b40c34f51f00888bec889d892fd098 Mon Sep 17 00:00:00 2001
From: Zhijie Hou <[email protected]>
Date: Tue, 9 Jun 2026 18:42:48 +0800
Subject: [PATCH v21 6/9] Track dependencies for streamed transactions
This commit allows tracking dependencies of streamed transactions.
Regarding the streaming=on case, dependency tracking is enabled while applying
spooled changes from files.
In the streaming=parallel case, dependency tracking is performed when the leader
sends changes to parallel workers. Apart from non-streamed transactions, the
leader waits for parallel workers till the assigned transactions are finished at
COMMIT/PREPARE/ABORT; thus, the XID of streamed transactions is not cached as
the lastly handled one. Also, streamed transactions are not recorded as
parallelized transactions because upcoming workers do not have to wait for them.
---
.../replication/logical/applyparallelworker.c | 8 ++-
src/backend/replication/logical/worker.c | 62 ++++++++++++++----
src/include/replication/worker_internal.h | 1 -
src/test/subscription/t/050_parallel_apply.pl | 63 +++++++++++++++++++
4 files changed, 120 insertions(+), 14 deletions(-)
diff --git a/src/backend/replication/logical/applyparallelworker.c b/src/backend/replication/logical/applyparallelworker.c
index 4a7f65345d8..6e478a41e73 100644
--- a/src/backend/replication/logical/applyparallelworker.c
+++ b/src/backend/replication/logical/applyparallelworker.c
@@ -165,7 +165,9 @@
* key) as another ongoing transaction (see handle_dependency_on_change for
* details). If so, the leader sends a list of dependent transaction IDs to the
* parallel worker, indicating that the parallel apply worker must wait for
- * these transactions to commit before proceeding.
+ * these transactions to commit before proceeding. If transactions are streamed
+ * but leader deciedes not to assign parallel apply workers, dependencies are
+ * verified when the transaction is committed.
*
* Tracking dependencies is necessary even when commit order is preserved.
* Consider two transactions: TX-1 (INSERT row 1) and TX-2 (DELETE row 1). If
@@ -1864,6 +1866,10 @@ pa_stream_abort(LogicalRepStreamAbortData *abort_data)
TransactionId xid = abort_data->xid;
TransactionId subxid = abort_data->subxid;
+ /* Streamed transactions won't be registered */
+ Assert(!dshash_find(parallelized_txns, &xid, false) &&
+ !dshash_find(parallelized_txns, &subxid, false));
+
/*
* Update origin state so we can restart streaming from correct position
* in case of crash.
diff --git a/src/backend/replication/logical/worker.c b/src/backend/replication/logical/worker.c
index 1582b7a6c58..0afba345713 100644
--- a/src/backend/replication/logical/worker.c
+++ b/src/backend/replication/logical/worker.c
@@ -1558,13 +1558,22 @@ handle_streamed_transaction(LogicalRepMsgType action, StringInfo s)
case TRANS_LEADER_SEND_TO_PARALLEL:
Assert(winfo);
+ /*
+ * Pass InvalidTransactionId to skip dependency recording for this
+ * change. Streaming transactions are assumed not to conflict with
+ * other transactions, so subsequent transactions do not need to
+ * wait for them to finish.
+ */
+ handle_dependency_on_change(action, s, InvalidTransactionId, winfo);
+
/*
* XXX The publisher side doesn't always send relation/type update
* messages after the streaming transaction, so also update the
* relation/type in leader apply worker. See function
* cleanup_rel_sync_cache.
*/
- if (pa_send_data(winfo, s->len, s->data))
+ if (!winfo->serialize_changes &&
+ pa_send_data(winfo, s->len, s->data))
return (action != LOGICAL_REP_MSG_RELATION &&
action != LOGICAL_REP_MSG_TYPE);
@@ -2695,6 +2704,13 @@ apply_handle_stream_prepare(StringInfo s)
apply_spooled_messages(MyLogicalRepWorker->stream_fileset,
prepare_data.xid, prepare_data.prepare_lsn);
+ /* Wait until the last transaction finishes */
+ if (TransactionIdIsValid(last_parallelized_remote_xid))
+ {
+ pa_wait_for_depended_transaction(last_parallelized_remote_xid);
+ last_parallelized_remote_xid = InvalidTransactionId;
+ }
+
/* Mark the transaction as prepared. */
apply_handle_prepare_internal(&prepare_data);
@@ -2718,6 +2734,12 @@ apply_handle_stream_prepare(StringInfo s)
case TRANS_LEADER_SEND_TO_PARALLEL:
Assert(winfo);
+ /*
+ * Build a dependency between this transaction and the lastly
+ * committed transaction to preserve the commit order.
+ */
+ build_dependency_with_last_committed_txn(winfo);
+
if (pa_send_data(winfo, s->len, s->data))
{
/* Finish processing the streaming transaction. */
@@ -2784,6 +2806,11 @@ apply_handle_stream_prepare(StringInfo s)
pgstat_report_stat(false);
+ /*
+ * No need to update the last_parallelized_remote_xid here because leader
+ * worker always wait until streamed transactions finish.
+ */
+
/*
* Process any tables that are being synchronized in parallel, as well as
* any newly added tables or sequences.
@@ -2897,17 +2924,8 @@ apply_handle_stream_start(StringInfo s)
set_apply_error_context_xact(stream_xid, InvalidXLogRecPtr);
- /*
- * Try to allocate a worker for the streaming transaction.
- *
- * TODO: Support assigning streaming transactions to parallel apply workers
- * even when non-streaming transactions are running and better dependency
- * handling between streaming and non-streaming transactions.
- */
- if (first_segment &&
- am_leader_apply_worker() &&
- (!TransactionIdIsValid(last_parallelized_remote_xid) ||
- pa_transaction_committed(last_parallelized_remote_xid)))
+ /* Try to allocate a worker for the streaming transaction. */
+ if (first_segment)
pa_allocate_worker(stream_xid, true);
apply_action = get_transaction_apply_action(stream_xid, &winfo);
@@ -3571,6 +3589,13 @@ apply_handle_stream_commit(StringInfo s)
apply_spooled_messages(MyLogicalRepWorker->stream_fileset, xid,
commit_data.commit_lsn);
+ /* Wait until the last transaction finishes */
+ if (TransactionIdIsValid(last_parallelized_remote_xid))
+ {
+ pa_wait_for_depended_transaction(last_parallelized_remote_xid);
+ last_parallelized_remote_xid = InvalidTransactionId;
+ }
+
apply_handle_commit_internal(&commit_data);
/* Unlink the files with serialized changes and subxact info. */
@@ -3582,6 +3607,19 @@ apply_handle_stream_commit(StringInfo s)
case TRANS_LEADER_SEND_TO_PARALLEL:
Assert(winfo);
+ /*
+ * Apart from non-streaming case, no need to mark this transaction
+ * as parallelized. Because the leader waits until the streamed
+ * transaction is committed thus commit ordering is always
+ * preserved.
+ */
+
+ /*
+ * Build a dependency between this transaction and the lastly
+ * committed transaction to preserve the commit order.
+ */
+ build_dependency_with_last_committed_txn(winfo);
+
if (pa_send_data(winfo, s->len, s->data))
{
/* Finish processing the streaming transaction. */
diff --git a/src/include/replication/worker_internal.h b/src/include/replication/worker_internal.h
index 02a165ccc0d..3dde899f465 100644
--- a/src/include/replication/worker_internal.h
+++ b/src/include/replication/worker_internal.h
@@ -370,7 +370,6 @@ extern void pa_switch_to_partial_serialize(ParallelApplyWorkerInfo *winfo,
extern void pa_set_xact_state(ParallelApplyWorkerShared *wshared,
ParallelTransState xact_state);
extern void pa_set_stream_apply_worker(ParallelApplyWorkerInfo *winfo);
-
extern void pa_start_subtrans(TransactionId current_xid,
TransactionId top_xid);
extern void pa_reset_subtrans(void);
diff --git a/src/test/subscription/t/050_parallel_apply.pl b/src/test/subscription/t/050_parallel_apply.pl
index dc25018b69a..a6343d57bc5 100644
--- a/src/test/subscription/t/050_parallel_apply.pl
+++ b/src/test/subscription/t/050_parallel_apply.pl
@@ -393,4 +393,67 @@ $result =
$node_subscriber->safe_psql('postgres', "SELECT count(1) FROM regress_tab");
is ($result, 20, 'inserts are replicated to subscriber');
+##################################################
+# Test that streaming transactions respect commit order preservation when
+# non-streaming transactions are being applied in parallel workers.
+##################################################
+
+$node_publisher->append_conf('postgresql.conf',
+ "logical_decoding_work_mem = 64kB");
+$node_publisher->reload;
+
+# Truncate the data for upcoming tests
+$node_publisher->safe_psql('postgres', "TRUNCATE TABLE regress_tab;");
+$node_publisher->wait_for_catchup('regress_sub');
+
+# Attach the injection_point again
+$node_subscriber->safe_psql('postgres',
+ "SELECT injection_points_attach('parallel-worker-before-commit','wait');"
+);
+
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO regress_tab VALUES (generate_series(71, 80), 'test');");
+
+# Wait until the parallel worker enters the injection point.
+$node_subscriber->wait_for_event('logical replication parallel worker',
+ 'parallel-worker-before-commit');
+
+# Run a transaction which would be streamed
+my $h = $node_publisher->background_psql('postgres', on_error_stop => 0);
+
+$offset = -s $node_subscriber->logfile;
+
+$h->query_safe(
+ q{
+BEGIN;
+UPDATE regress_tab SET value = 'streamed-updated' WHERE id BETWEEN 71 AND 80;
+INSERT INTO regress_tab VALUES (generate_series(100, 5100), 'streamed');
+});
+
+# Verify the dependency is detected for the delete
+$str = $node_subscriber->wait_for_log(qr/found conflicting replica identity change on table [1-9][0-9]+ from ([1-9][0-9]+)/, $offset);
+$xid = $str =~ /found conflicting replica identity change on table [1-9][0-9]+ from ([1-9][0-9]+)/;
+
+# Verify the parallel worker waits for the same transaction
+$node_subscriber->wait_for_log(qr/wait for depended xid $xid/, $offset);
+
+ok(1, "replica identity dependency from streamed txn detected for parallel apply");
+
+# Wakeup the parallel worker
+$node_subscriber->safe_psql('postgres', qq[
+ SELECT injection_points_detach('parallel-worker-before-commit');
+ SELECT injection_points_wakeup('parallel-worker-before-commit');
+]);
+
+# Verify the streamed transaction can be applied
+$node_subscriber->wait_for_log(qr/finish waiting for depended xid $xid/, $offset);
+
+$h->query_safe("COMMIT;");
+
+$node_publisher->wait_for_catchup('regress_sub');
+
+$result =
+ $node_subscriber->safe_psql('postgres', "SELECT count(1) FROM regress_tab");
+is ($result, 5011, 'inserts are replicated to subscriber');
+
done_testing();
--
2.43.0
[application/octet-stream] v21-0005-support-2PC.patch (13.8K, ../TY4PR01MB177180394D8E9C6D263E117E494FE2@TY4PR01MB17718.jpnprd01.prod.outlook.com/7-v21-0005-support-2PC.patch)
download | inline diff:
From 92d467d086d11d8953664317645c51e249aacbcb Mon Sep 17 00:00:00 2001
From: Zhijie Hou <[email protected]>
Date: Wed, 8 Jul 2026 16:07:32 +0800
Subject: [PATCH v21 5/9] support 2PC
This patch allows the PREPARE transaction to be applied in parallel. Parallel
apply workers are assigned to a transaction when BEGIN_PREPARE is received. This
part and the dependency-waiting mechanism are the same as a normal transaction.
A parallel worker can be freed after it handles a PREPARE message. The prepared
transaction can be deleted from parallelized_txns at that time; the upcoming
transactions will wait until then.
The leader apply worker resolves COMMIT PREPARED/ROLLBACK PREPARED. Since it can
be serialized automatically, it does not add the transaction to parallelized_txns.
---
src/backend/replication/logical/worker.c | 218 +++++++++++++++---
src/test/subscription/t/050_parallel_apply.pl | 74 +++++-
2 files changed, 263 insertions(+), 29 deletions(-)
diff --git a/src/backend/replication/logical/worker.c b/src/backend/replication/logical/worker.c
index b5013cbfc00..1582b7a6c58 100644
--- a/src/backend/replication/logical/worker.c
+++ b/src/backend/replication/logical/worker.c
@@ -810,6 +810,14 @@ delete_replica_identity_entries_for_txns(List *committed_xids)
if (!committed_xids)
return;
+ /*
+ * Skip if the replica_identity_table is not initialized yet. This can
+ * happen if the empty transaction was replicated and a parallel apply
+ * worker was launched. See comments in apply_handle_prepare().
+ */
+ if (!replica_identity_table)
+ return;
+
replica_identity_start_iterate(replica_identity_table, &i);
while ((rientry = replica_identity_iterate(replica_identity_table, &i)) != NULL)
{
@@ -2244,6 +2252,8 @@ static void
apply_handle_begin_prepare(StringInfo s)
{
LogicalRepPreparedTxnData begin_data;
+ ParallelApplyWorkerInfo *winfo;
+ TransApplyAction apply_action;
/* Tablesync should never receive prepare. */
if (am_tablesync_worker())
@@ -2255,12 +2265,54 @@ apply_handle_begin_prepare(StringInfo s)
Assert(!TransactionIdIsValid(stream_xid));
logicalrep_read_begin_prepare(s, &begin_data);
- set_apply_error_context_xact(begin_data.xid, begin_data.prepare_lsn);
+
+ remote_xid = begin_data.xid;
+
+ set_apply_error_context_xact(remote_xid, begin_data.prepare_lsn);
remote_final_lsn = begin_data.prepare_lsn;
maybe_start_skipping_changes(begin_data.prepare_lsn);
+ pa_allocate_worker(remote_xid, false);
+
+ apply_action = get_transaction_apply_action(remote_xid, &winfo);
+
+ elog(DEBUG1, "new remote_xid %u", remote_xid);
+ switch (apply_action)
+ {
+ case TRANS_LEADER_APPLY:
+ break;
+
+ case TRANS_LEADER_SEND_TO_PARALLEL:
+ Assert(winfo);
+
+ if (pa_send_data(winfo, s->len, s->data))
+ {
+ pa_set_stream_apply_worker(winfo);
+ break;
+ }
+
+ /*
+ * TODO: Support switching to PARTIAL_SERIALIZE mode when the send
+ * buffer becomes full.
+ */
+ ereport(ERROR,
+ errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
+ errmsg("could not send data to the logical replication parallel apply worker"));
+ break;
+
+ case TRANS_PARALLEL_APPLY:
+ /* Hold the lock until the end of the transaction. */
+ pa_lock_transaction(MyParallelShared->xid, AccessExclusiveLock);
+ pa_set_xact_state(MyParallelShared, PARALLEL_TRANS_STARTED);
+ break;
+
+ default:
+ elog(ERROR, "unexpected apply action: %d", (int) apply_action);
+ break;
+ }
+
in_remote_transaction = true;
pgstat_report_activity(STATE_RUNNING, NULL);
@@ -2310,6 +2362,8 @@ static void
apply_handle_prepare(StringInfo s)
{
LogicalRepPreparedTxnData prepare_data;
+ ParallelApplyWorkerInfo *winfo;
+ TransApplyAction apply_action;
logicalrep_read_prepare(s, &prepare_data);
@@ -2320,36 +2374,131 @@ apply_handle_prepare(StringInfo s)
LSN_FORMAT_ARGS(prepare_data.prepare_lsn),
LSN_FORMAT_ARGS(remote_final_lsn))));
- /*
- * Unlike commit, here, we always prepare the transaction even though no
- * change has happened in this transaction or all changes are skipped. It
- * is done this way because at commit prepared time, we won't know whether
- * we have skipped preparing a transaction because of those reasons.
- *
- * XXX, We can optimize such that at commit prepared time, we first check
- * whether we have prepared the transaction or not but that doesn't seem
- * worthwhile because such cases shouldn't be common.
- */
- begin_replication_step();
+ apply_action = get_transaction_apply_action(remote_xid, &winfo);
- apply_handle_prepare_internal(&prepare_data);
+ switch (apply_action)
+ {
+ case TRANS_LEADER_APPLY:
- end_replication_step();
- CommitTransactionCommand();
- pgstat_report_stat(false);
+ /*
+ * Unlike commit, here, we always prepare the transaction even
+ * though no change has happened in this transaction or all
+ * changes are skipped. It is done this way because at commit
+ * prepared time, we won't know whether we have skipped preparing
+ * a transaction because of those reasons.
+ *
+ * XXX, We can optimize such that at commit prepared time, we
+ * first check whether we have prepared the transaction or not but
+ * that doesn't seem worthwhile because such cases shouldn't be
+ * common.
+ */
+ begin_replication_step();
- /*
- * It is okay not to set the local_end LSN for the prepare because we
- * always flush the prepare record. So, we can send the acknowledgment of
- * the remote_end LSN as soon as prepare is finished.
- *
- * XXX For the sake of consistency with commit, we could have set it with
- * the LSN of prepare but as of now we don't track that value similar to
- * XactLastCommitEnd, and adding it for this purpose doesn't seems worth
- * it.
- */
- store_flush_position(prepare_data.end_lsn, InvalidXLogRecPtr,
- InvalidTransactionId);
+ /* Wait until the last transaction finishes */
+ if (TransactionIdIsValid(last_parallelized_remote_xid))
+ {
+ pa_wait_for_depended_transaction(last_parallelized_remote_xid);
+ last_parallelized_remote_xid = InvalidTransactionId;
+ }
+
+ apply_handle_prepare_internal(&prepare_data);
+
+ end_replication_step();
+ CommitTransactionCommand();
+ pgstat_report_stat(false);
+
+ /*
+ * It is okay not to set the local_end LSN for the prepare because
+ * we always flush the prepare record. So, we can send the
+ * acknowledgment of the remote_end LSN as soon as prepare is
+ * finished.
+ *
+ * XXX For the sake of consistency with commit, we could have set
+ * it with the LSN of prepare but as of now we don't track that
+ * value similar to XactLastCommitEnd, and adding it for this
+ * purpose doesn't seems worth it.
+ */
+ store_flush_position(prepare_data.end_lsn, InvalidXLogRecPtr,
+ InvalidTransactionId);
+
+ break;
+
+ case TRANS_LEADER_SEND_TO_PARALLEL:
+ Assert(winfo);
+
+ /*
+ * Mark this transaction as parallelized. This ensures that
+ * upcoming transactions wait until this transaction is committed.
+ */
+ pa_add_parallelized_transaction(remote_xid);
+
+ /*
+ * Build a dependency between this transaction and the lastly
+ * committed transaction to preserve the commit order.
+ */
+ build_dependency_with_last_committed_txn(winfo);
+
+ if (pa_send_data(winfo, s->len, s->data))
+ {
+ /* Cache the remote_xid */
+ last_parallelized_remote_xid = remote_xid;
+
+ /* Finish processing the transaction. */
+ pa_xact_finish(winfo, prepare_data.end_lsn);
+ break;
+ }
+
+ /*
+ * TODO: Support switching to PARTIAL_SERIALIZE mode when the send
+ * buffer becomes full.
+ */
+ ereport(ERROR,
+ errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
+ errmsg("could not send data to the logical replication parallel apply worker"));
+ break;
+
+ case TRANS_PARALLEL_APPLY:
+
+ /*
+ * If the parallel apply worker is applying spooled messages then
+ * close the file before committing.
+ */
+ if (stream_fd)
+ stream_close_file();
+
+ begin_replication_step();
+
+ INJECTION_POINT("parallel-worker-before-prepare", NULL);
+
+ /* Mark the transaction as prepared. */
+ apply_handle_prepare_internal(&prepare_data);
+
+ end_replication_step();
+
+ CommitTransactionCommand();
+ pgstat_report_stat(false);
+
+ store_flush_position(prepare_data.end_lsn, InvalidXLogRecPtr,
+ InvalidTransactionId);
+
+ /*
+ * It is okay not to set the local_end LSN for the prepare because
+ * we always flush the prepare record. See apply_handle_prepare.
+ */
+ MyParallelShared->last_commit_end = InvalidXLogRecPtr;
+ pa_commit_transaction();
+
+ pa_unlock_transaction(MyParallelShared->xid, AccessExclusiveLock);
+
+ pa_reset_subtrans();
+ break;
+
+ default:
+ elog(ERROR, "unexpected apply action: %d", (int) apply_action);
+ break;
+ }
+
+ remote_xid = InvalidTransactionId;
in_remote_transaction = false;
@@ -2397,6 +2546,12 @@ apply_handle_commit_prepared(StringInfo s)
/* There is no transaction when COMMIT PREPARED is called */
begin_replication_step();
+ if (TransactionIdIsValid(last_parallelized_remote_xid))
+ {
+ pa_wait_for_depended_transaction(last_parallelized_remote_xid);
+ last_parallelized_remote_xid = InvalidTransactionId;
+ }
+
/*
* Update origin state so we can restart streaming from correct position
* in case of crash.
@@ -2465,6 +2620,13 @@ apply_handle_rollback_prepared(StringInfo s)
/* There is no transaction when ABORT/ROLLBACK PREPARED is called */
begin_replication_step();
+
+ if (TransactionIdIsValid(last_parallelized_remote_xid))
+ {
+ pa_wait_for_depended_transaction(last_parallelized_remote_xid);
+ last_parallelized_remote_xid = InvalidTransactionId;
+ }
+
FinishPreparedTransaction(gid, false);
end_replication_step();
CommitTransactionCommand();
diff --git a/src/test/subscription/t/050_parallel_apply.pl b/src/test/subscription/t/050_parallel_apply.pl
index 9d6026470d7..dc25018b69a 100644
--- a/src/test/subscription/t/050_parallel_apply.pl
+++ b/src/test/subscription/t/050_parallel_apply.pl
@@ -19,6 +19,8 @@ if ($ENV{enable_injection_points} ne 'yes')
# Initialize publisher node
my $node_publisher = PostgreSQL::Test::Cluster->new('publisher');
$node_publisher->init(allows_streaming => 'logical');
+$node_publisher->append_conf('postgresql.conf',
+ "max_prepared_transactions = 10");
$node_publisher->start;
# Create tables and insert initial data
@@ -48,7 +50,8 @@ my $node_subscriber = PostgreSQL::Test::Cluster->new('subscriber');
$node_subscriber->init;
$node_subscriber->append_conf('postgresql.conf', "log_min_messages = debug1");
$node_subscriber->append_conf('postgresql.conf',
- "max_logical_replication_workers = 10");
+ "max_logical_replication_workers = 10
+ max_prepared_transactions = 10");
$node_subscriber->start;
# Check if the extension injection_points is available, as it may be
@@ -321,4 +324,73 @@ $result =
$node_subscriber->safe_psql('postgres', "SELECT count(1) FROM regress_tab");
is ($result, 1, 'inserts are replicated to subscriber');
+##################################################
+# Test that the prepared transaction can be applied in a parallel apply worker,
+# and that the commit order preservation work correctly.
+##################################################
+
+# Truncate the data for upcoming tests
+$node_publisher->safe_psql('postgres', "TRUNCATE TABLE regress_tab;");
+$node_publisher->wait_for_catchup('regress_sub');
+
+$node_subscriber->safe_psql('postgres',
+ "ALTER SUBSCRIPTION regress_sub DISABLE;");
+$node_subscriber->poll_query_until('postgres',
+ "SELECT count(*) = 0 FROM pg_stat_activity WHERE backend_type = 'logical replication apply worker'"
+);
+$node_subscriber->safe_psql(
+ 'postgres', "
+ ALTER SUBSCRIPTION regress_sub SET (two_phase = on);
+ ALTER SUBSCRIPTION regress_sub ENABLE;");
+
+$result = $node_subscriber->safe_psql('postgres',
+ "SELECT count(1) FROM pg_stat_activity WHERE backend_type = 'logical replication parallel worker'");
+is($result, '0', "no parallel apply workers exist after restart");
+
+# Attach an injection_point. Parallel workers would wait before the prepare
+$node_subscriber->safe_psql('postgres',
+ "SELECT injection_points_attach('parallel-worker-before-prepare','wait');"
+);
+
+# PREPARE a transaction on publisher. It would be handled by a parallel apply
+# worker.
+$node_publisher->safe_psql('postgres', qq[
+ BEGIN;
+ INSERT INTO regress_tab VALUES (generate_series(51, 60), 'prepare');
+ PREPARE TRANSACTION 'regress_prepare';
+]);
+
+# Wait until the parallel worker enters the injection point.
+$node_subscriber->wait_for_event('logical replication parallel worker',
+ 'parallel-worker-before-prepare');
+
+$offset = -s $node_subscriber->logfile;
+
+# Insert tuples on publisher again. This transaction waits for the prepared
+# transaction
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO regress_tab VALUES (generate_series(61, 70), 'test');");
+
+# Verify the parallel worker waits for the transaction
+$str = $node_subscriber->wait_for_log(qr/wait for depended xid ([1-9][0-9]+)/, $offset);
+$xid = $str =~ /wait for depended xid ([1-9][0-9]+)/;
+
+ok(1, "commit order dependency from prepared transaction detected for parallel apply");
+
+# Wakeup the parallel worker
+$node_subscriber->safe_psql('postgres', qq[
+ SELECT injection_points_detach('parallel-worker-before-prepare');
+ SELECT injection_points_wakeup('parallel-worker-before-prepare');
+]);
+
+$node_subscriber->wait_for_log(qr/finish waiting for depended xid $xid/, $offset);
+
+# COMMIT the prepared transaction. It is always handled by the leader
+$node_publisher->safe_psql('postgres', "COMMIT PREPARED 'regress_prepare';");
+$node_publisher->wait_for_catchup('regress_sub');
+
+$result =
+ $node_subscriber->safe_psql('postgres', "SELECT count(1) FROM regress_tab");
+is ($result, 20, 'inserts are replicated to subscriber');
+
done_testing();
--
2.43.0
[application/octet-stream] v21-0003-Introduce-a-local-hash-table-to-store-replica-id.patch (35.9K, ../TY4PR01MB177180394D8E9C6D263E117E494FE2@TY4PR01MB17718.jpnprd01.prod.outlook.com/8-v21-0003-Introduce-a-local-hash-table-to-store-replica-id.patch)
download | inline diff:
From eaa42b2a5ff7585284e694437f2c2b40eccf35c9 Mon Sep 17 00:00:00 2001
From: Zhijie Hou <[email protected]>
Date: Mon, 15 Jun 2026 18:51:14 +0800
Subject: [PATCH v21 3/9] Introduce a local hash table to store replica
identities
This local hash table on the leader is used for detecting dependencies between
transactions.
The hash contains the Replica Identity (RI) as a key and the remote XID that
modified the corresponding tuple. The hash entries are inserted when the leader
finds an RI from a replication message, and are removed in the
following cases:
1) When handling a subsequent change that modifies the same row, and the
stored transaction has already committed.
2) When collecting flush progress of remote transactions, and the stored
transaction is found to have committed.
3) When the number of entries exceeds REPLICA_IDENTITY_CLEANUP_THRESHOLD
and the stored transaction has already committed.
When the leader sends replication changes to parallel workers, it checks whether
aother transaction has already used the RI associated with the change. If so,
the leader marks the current transaction as dependent on that transaction and
notifies parallel workers via PA_MSG_XACT_DEPENDENCY to wait for that
transaction to finish.
Author: Zhijie Hou <[email protected]>
Author: Hayato Kuroda <[email protected]>
---
.../replication/logical/applyparallelworker.c | 131 +++-
src/backend/replication/logical/relation.c | 24 +
src/backend/replication/logical/worker.c | 739 +++++++++++++++++-
src/include/replication/logicalrelation.h | 9 +
src/include/replication/worker_internal.h | 6 +-
src/test/subscription/meson.build | 1 +
src/tools/pgindent/typedefs.list | 3 +
7 files changed, 908 insertions(+), 5 deletions(-)
diff --git a/src/backend/replication/logical/applyparallelworker.c b/src/backend/replication/logical/applyparallelworker.c
index efde9181a34..30d31e02ac1 100644
--- a/src/backend/replication/logical/applyparallelworker.c
+++ b/src/backend/replication/logical/applyparallelworker.c
@@ -213,17 +213,41 @@
#define PARALLEL_APPLY_LOCK_XACT 1
/*
- * Hash table entry to map xid to the parallel apply worker state.
+ * Hash table entry of ParallelApplyTxnHash to map xid to the parallel apply
+ * worker state.
*/
typedef struct ParallelApplyWorkerEntry
{
- TransactionId xid; /* Hash key -- must be first */
+ TransactionId xid; /* Remote transaction ID (hash key) ---
+ * must be first */
+
+ /*
+ * The parallel apply worker assigned for applying the transaction. It is
+ * NULL if the worker has already finished the transaction and the leader
+ * has collected its local end LSN.
+ */
ParallelApplyWorkerInfo *winfo;
+
+ /*
+ * The local end LSN of this transaction applied by the parallel apply
+ * worker.
+ *
+ * The leader will initialize this value by reading the last_commit_end from
+ * the parallel apply worker shared memory before reusing the worker for
+ * another transaction (see pa_get_last_commit_end). An invalid value
+ * indicates the worker has not finished applying the transaction or there
+ * is no change applied by the worker for this transaction.
+ */
+ XLogRecPtr local_end;
} ParallelApplyWorkerEntry;
/*
* A hash table used to cache the state of streaming transactions being applied
* by the parallel apply workers. Entries are of type ParallelApplyWorkerEntry.
+ *
+ * The leader apply worker adds an entry when a transaction starts being
+ * applied, and removes it after collecting the transaction's local end LSN from
+ * the parallel worker.
*/
static HTAB *ParallelApplyTxnHash = NULL;
@@ -237,6 +261,10 @@ typedef struct ParallelizedTxnEntry
* A hash table used to track the parallelized remote transactions that could be
* depended on by other transactions. Entries are of type ParallelizedTxnEntry.
*
+ * The leader apply worker adds an entry before sending the COMMIT record of a
+ * transaction to a parallel apply worker. The entry is removed by that worker
+ * immediately after committing the transaction.
+ *
* dshash is used to enable dynamic shared memory allocation based on the number
* of transactions being applied in parallel.
*/
@@ -510,7 +538,7 @@ pa_launch_parallel_worker(void)
* streaming changes.
*/
void
-pa_allocate_worker(TransactionId xid)
+pa_allocate_worker(TransactionId xid, bool stream_txn)
{
bool found;
ParallelApplyWorkerInfo *winfo = NULL;
@@ -551,7 +579,9 @@ pa_allocate_worker(TransactionId xid)
winfo->in_use = true;
winfo->serialize_changes = false;
+ winfo->stream_txn = stream_txn;
entry->winfo = winfo;
+ entry->local_end = InvalidXLogRecPtr;
}
/*
@@ -748,6 +778,72 @@ pa_process_spooled_messages_if_required(void)
return true;
}
+/*
+ * Get the local end LSN for a transaction applied by a parallel apply worker.
+ *
+ * If the parallel apply worker did not write any changes during the transaction
+ * application due to situations like update/delete_missing or a before trigger,
+ * the *skipped_write will be set to true.
+ *
+ * Once the LSN is retrieved, the transaction entry is removed from
+ * ParallelApplyTxnHash.
+ */
+XLogRecPtr
+pa_get_last_commit_end(TransactionId xid, bool *skipped_write)
+{
+ bool found;
+ ParallelApplyWorkerEntry *entry;
+ ParallelApplyWorkerInfo *winfo;
+
+ Assert(TransactionIdIsValid(xid));
+ Assert(ParallelApplyTxnHash);
+
+ if (skipped_write)
+ *skipped_write = false;
+
+ /* Find an entry for the requested transaction. */
+ entry = hash_search(ParallelApplyTxnHash, &xid, HASH_FIND, &found);
+
+ if (!found)
+ return InvalidXLogRecPtr;
+
+ /*
+ * If worker info is NULL, it indicates that the worker has already
+ * finished handling the transaction. Consequently, the local end LSN has
+ * already been collected and saved in entry->local_end.
+ */
+ winfo = entry->winfo;
+ if (winfo == NULL)
+ {
+ if (!hash_search(ParallelApplyTxnHash, &xid, HASH_REMOVE, NULL))
+ elog(ERROR, "hash table corrupted");
+
+ if (skipped_write)
+ *skipped_write = XLogRecPtrIsInvalid(entry->local_end);
+
+ return entry->local_end;
+ }
+
+ /* Return InvalidXLogRecPtr if the transaction is still in progress */
+ if (pa_get_xact_state(winfo->shared) != PARALLEL_TRANS_FINISHED)
+ return InvalidXLogRecPtr;
+
+ /* Collect the local end LSN from the worker's shared memory area */
+ entry->local_end = winfo->shared->last_commit_end;
+ entry->winfo = NULL;
+
+ if (skipped_write)
+ *skipped_write = XLogRecPtrIsInvalid(entry->local_end);
+
+ elog(DEBUG1, "store local commit %X/%X end to txn entry: %u",
+ LSN_FORMAT_ARGS(entry->local_end), xid);
+
+ if (!hash_search(ParallelApplyTxnHash, &xid, HASH_REMOVE, NULL))
+ elog(ERROR, "hash table corrupted");
+
+ return entry->local_end;
+}
+
/*
* Interrupt handler for main loop of parallel apply worker.
*/
@@ -1795,6 +1891,35 @@ pa_xact_finish(ParallelApplyWorkerInfo *winfo, XLogRecPtr remote_lsn)
pa_free_worker(winfo);
}
+/*
+ * Check if the given transaction has been committed. This can only be used by
+ * the leader apply worker.
+ *
+ * Returns true if the transaction is committed, false if it is still being
+ * applied in parallel.
+ */
+bool
+pa_transaction_committed(TransactionId xid)
+{
+ bool found;
+ ParallelApplyWorkerEntry *entry;
+
+ Assert(am_leader_apply_worker());
+ Assert(TransactionIdIsValid(xid));
+ Assert(ParallelApplyTxnHash);
+
+ /* Find an entry for the requested transaction */
+ entry = hash_search(ParallelApplyTxnHash, &xid, HASH_FIND, &found);
+
+ if (!found)
+ return true;
+
+ if (!entry->winfo)
+ return true;
+
+ return pa_get_xact_state(entry->winfo->shared) == PARALLEL_TRANS_FINISHED;
+}
+
/*
* Attach to the shared hash table for parallelized transactions.
*/
diff --git a/src/backend/replication/logical/relation.c b/src/backend/replication/logical/relation.c
index 296cbaede30..774e048c263 100644
--- a/src/backend/replication/logical/relation.c
+++ b/src/backend/replication/logical/relation.c
@@ -959,3 +959,27 @@ FindLogicalRepLocalIndex(Relation localrel, LogicalRepRelation *remoterel,
return InvalidOid;
}
+
+/*
+ * Get the LogicalRepRelMapEntry corresponding to the given relid without
+ * opening the local relation.
+ */
+LogicalRepRelMapEntry *
+logicalrep_get_relentry(LogicalRepRelId remoteid)
+{
+ LogicalRepRelMapEntry *entry;
+ bool found;
+
+ if (LogicalRepRelMap == NULL)
+ logicalrep_relmap_init();
+
+ /* Search for existing entry. */
+ entry = hash_search(LogicalRepRelMap, (void *) &remoteid,
+ HASH_FIND, &found);
+
+ if (!found)
+ elog(DEBUG1, "no relation map entry for remote relation ID %u",
+ remoteid);
+
+ return entry;
+}
diff --git a/src/backend/replication/logical/worker.c b/src/backend/replication/logical/worker.c
index bd53034cf69..3a1423fbe58 100644
--- a/src/backend/replication/logical/worker.c
+++ b/src/backend/replication/logical/worker.c
@@ -303,11 +303,35 @@
#define NAPTIME_PER_CYCLE 1000 /* max sleep time between cycles (1s) */
+/*
+ * Struct for tracking the progress of flushing the transaction received from
+ * the publisher.
+ *
+ * A list of these structs (lsn_mapping) is maintained in the apply worker, each
+ * representing a transaction that is being flushed. The entries are removed
+ * from the list when the transaction is fully flushed (see get_flush_position).
+ */
typedef struct FlushPosition
{
dlist_node node;
+
+ /*
+ * The end of commit record applied by the worker, and the corresponding end
+ * of commit record on the publisher.
+ *
+ * For transactions assigned to a parallel apply worker, the local_end is
+ * not immediately available. The leader will update the local_end when it
+ * confirms that the parallel apply worker has finished applying the
+ * transaction (see the usage of pa_get_last_commit_end).
+ */
XLogRecPtr local_end;
XLogRecPtr remote_end;
+
+ /*
+ * The remote transaction ID. This should be set to a valid ID only when the
+ * transaction is assigned to a parallel apply worker.
+ */
+ TransactionId pa_remote_xid;
} FlushPosition;
static dlist_head lsn_mapping = DLIST_STATIC_INIT(lsn_mapping);
@@ -476,6 +500,8 @@ ErrorContextCallback *apply_error_context_stack = NULL;
MemoryContext ApplyMessageContext = NULL;
MemoryContext ApplyContext = NULL;
+static MemoryContext ParallelApplyContext = NULL;
+
/* per stream context for streaming transactions */
static MemoryContext LogicalStreamingContext = NULL;
@@ -549,6 +575,65 @@ typedef struct ApplySubXactData
static ApplySubXactData subxact_data = {0, 0, InvalidTransactionId, NULL};
+/* Hash table key for replica_identity_table */
+typedef struct ReplicaIdentityKey
+{
+ Oid relid;
+ LogicalRepTupleData *data;
+} ReplicaIdentityKey;
+
+/* Hash table entry in replica_identity_table */
+typedef struct ReplicaIdentityEntry
+{
+ ReplicaIdentityKey *keydata;
+ TransactionId remote_xid;
+
+ /* needed for simplehash */
+ uint32 hash;
+ char status;
+} ReplicaIdentityEntry;
+
+#include "common/hashfn.h"
+
+static uint32 hash_replica_identity(ReplicaIdentityKey *key);
+static bool hash_replica_identity_compare(ReplicaIdentityKey *a,
+ ReplicaIdentityKey *b);
+
+/* Define parameters for replica identity hash table code generation. */
+#define SH_PREFIX replica_identity
+#define SH_ELEMENT_TYPE ReplicaIdentityEntry
+#define SH_KEY_TYPE ReplicaIdentityKey *
+#define SH_KEY keydata
+#define SH_HASH_KEY(tb, key) hash_replica_identity(key)
+#define SH_EQUAL(tb, a, b) hash_replica_identity_compare(a, b)
+#define SH_STORE_HASH
+#define SH_GET_HASH(tb, a) (a)->hash
+#define SH_SCOPE static inline
+#define SH_DECLARE
+#define SH_DEFINE
+#include "lib/simplehash.h"
+
+#define REPLICA_IDENTITY_INITIAL_SIZE 128
+#define REPLICA_IDENTITY_CLEANUP_THRESHOLD 1024
+
+/*
+ * Hash table storing replica identity values for changes being applied in
+ * parallel, along with the latest transaction that modified each row.
+ *
+ * Entries are removed in the following cases:
+ *
+ * 1) When handling a subsequent change that modifies the same row, and the
+ * stored transaction has already committed.
+ * 2) When collecting flush progress of remote transactions, and the stored
+ * transaction is found to have committed.
+ * 3) When the number of entries exceeds REPLICA_IDENTITY_CLEANUP_THRESHOLD
+ * and the stored transaction has already committed.
+ *
+ * We use simplehash for its efficiency (see comments atop of simplehash.h), as
+ * this hash table is accessed for every change in a transaction.
+ */
+static replica_identity_hash *replica_identity_table = NULL;
+
static inline void subxact_filename(char *path, Oid subid, TransactionId xid);
static inline void changes_filename(char *path, Oid subid, TransactionId xid);
@@ -636,6 +721,631 @@ static void set_wal_receiver_timeout(void);
static void on_exit_clear_xact_state(int code, Datum arg);
+static void send_internal_dependencies(ParallelApplyWorkerInfo *winfo,
+ List *depends_on_xids);
+
+/*
+ * Compute the hash value for entries in the replica_identity_table.
+ */
+static uint32
+hash_replica_identity(ReplicaIdentityKey *key)
+{
+ int i;
+ uint32 hashkey = 0;
+
+ hashkey = hash_combine(hashkey, hash_uint32(key->relid));
+
+ for (i = 0; i < key->data->ncols; i++)
+ {
+ uint32 hkey;
+
+ if (key->data->colstatus[i] == LOGICALREP_COLUMN_NULL)
+ continue;
+
+ hkey = hash_any((const unsigned char *) key->data->colvalues[i].data,
+ key->data->colvalues[i].len);
+ hashkey = hash_combine(hashkey, hkey);
+ }
+
+ return hashkey;
+}
+
+/*
+ * Compare two entries in the replica_identity_table.
+ */
+static bool
+hash_replica_identity_compare(ReplicaIdentityKey *a, ReplicaIdentityKey *b)
+{
+ if (a->relid != b->relid ||
+ a->data->ncols != b->data->ncols)
+ return false;
+
+ for (int i = 0; i < a->data->ncols; i++)
+ {
+ if (a->data->colstatus[i] != b->data->colstatus[i])
+ return false;
+
+ if (a->data->colvalues[i].len != b->data->colvalues[i].len)
+ return false;
+
+ if (strcmp(a->data->colvalues[i].data, b->data->colvalues[i].data))
+ return false;
+
+ elog(DEBUG1, "conflicting key %s", a->data->colvalues[i].data);
+ }
+
+ return true;
+}
+
+/*
+ * Free resources associated with a replica identity key.
+ */
+static void
+free_replica_identity_key(ReplicaIdentityKey *key)
+{
+ Assert(key);
+
+ pfree(key->data->colvalues);
+ pfree(key->data->colstatus);
+ pfree(key->data);
+ pfree(key);
+}
+
+/*
+ * Clean up hash table entries associated with the given transaction IDs.
+ */
+static void
+delete_replica_identity_entries_for_txns(List *committed_xids)
+{
+ replica_identity_iterator i;
+ ReplicaIdentityEntry *rientry;
+
+ if (!committed_xids)
+ return;
+
+ replica_identity_start_iterate(replica_identity_table, &i);
+ while ((rientry = replica_identity_iterate(replica_identity_table, &i)) != NULL)
+ {
+ if (!list_member_xid(committed_xids, rientry->remote_xid))
+ continue;
+
+ /* Clean up the hash entry for committed transaction */
+ free_replica_identity_key(rientry->keydata);
+ replica_identity_delete_item(replica_identity_table, rientry);
+ }
+}
+
+/*
+ * Check committed transactions and clean up corresponding entries in the hash
+ * table.
+ *
+ * Also update the local_end for transactions assigned to parallel apply workers
+ * if not already set.
+ */
+static void
+cleanup_committed_replica_identity_entries(void)
+{
+ dlist_mutable_iter iter;
+ List *committed_xids = NIL;
+
+ dlist_foreach_modify(iter, &lsn_mapping)
+ {
+ FlushPosition *pos =
+ dlist_container(FlushPosition, node, iter.cur);
+ bool skipped_write;
+
+ /*
+ * Skip if the transaction was not assigned to a parallel worker, or if
+ * we have already collected its local end LSN.
+ */
+ if (!TransactionIdIsValid(pos->pa_remote_xid) ||
+ XLogRecPtrIsValid(pos->local_end))
+ continue;
+
+ pos->local_end = pa_get_last_commit_end(pos->pa_remote_xid,
+ &skipped_write);
+
+ elog(DEBUG1,
+ "got commit end from parallel apply worker, "
+ "txn: %u, remote_end %X/%X, local_end %X/%X",
+ pos->pa_remote_xid, LSN_FORMAT_ARGS(pos->remote_end),
+ LSN_FORMAT_ARGS(pos->local_end));
+
+ /* Skip if the transaction is still being applied */
+ if (!skipped_write && !XLogRecPtrIsValid(pos->local_end))
+ continue;
+
+ committed_xids = lappend_xid(committed_xids, pos->pa_remote_xid);
+ }
+
+ /* cleanup the entries for committed transactions */
+ delete_replica_identity_entries_for_txns(committed_xids);
+}
+
+/*
+ * Check whether the given relation entry contains an uncommitted transaction
+ * that affected the whole table.
+ */
+static bool
+has_active_rel_dependency(LogicalRepRelMapEntry *relentry)
+{
+ if (!TransactionIdIsValid(relentry->last_depended_xid))
+ return false;
+
+ if (pa_transaction_committed(relentry->last_depended_xid))
+ {
+ relentry->last_depended_xid = InvalidTransactionId;
+ return false;
+ }
+
+ return true;
+}
+
+/*
+ * Check whether the given key dependency entry contains an uncommitted
+ * transaction. If cleanup is true, also clean up the entry if the transaction
+ * has already committed.
+ */
+static bool
+has_active_key_dependency(ReplicaIdentityEntry *rientry, bool cleanup)
+{
+ Assert(TransactionIdIsValid(rientry->remote_xid));
+
+ if (pa_transaction_committed(rientry->remote_xid))
+ {
+ if (cleanup)
+ {
+ free_replica_identity_key(rientry->keydata);
+ replica_identity_delete_item(replica_identity_table, rientry);
+ }
+
+ return false;
+ }
+
+ return true;
+}
+
+/*
+ * Append the given transaction ID to the dependency list if it is not already
+ * present.
+ */
+static void
+append_xid_dependency(TransactionId xid, List **depends_on_xids)
+{
+ Assert(TransactionIdIsValid(xid));
+
+ if (list_member_xid(*depends_on_xids, xid))
+ return;
+
+ *depends_on_xids = lappend_xid(*depends_on_xids, xid);
+}
+
+/*
+ * Check if any of the key columns have NULL values.
+ */
+static bool
+has_null_key_values(LogicalRepTupleData *data, Bitmapset *indexkeys)
+{
+ for (int i = 0; i < data->ncols; i++)
+ {
+ if (bms_is_member(i, indexkeys) &&
+ data->colstatus[i] == LOGICALREP_COLUMN_NULL)
+ return true;
+ }
+
+ return false;
+}
+
+/*
+ * Check for dependencies on preceding transactions that modify the same key as
+ * the given tuple. Returns the dependent transactions in 'depends_on_xids'.
+ *
+ * Additionally, if new_depended_xid is valid, record the current change and the
+ * transaction as a new dependency for the replica identity key modification,
+ * allowing subsequent transactions that modify the same key to be dependent on
+ * it.
+ */
+static void
+check_and_record_ri_dependency(Oid relid, LogicalRepTupleData *original_data,
+ TransactionId new_depended_xid,
+ List **depends_on_xids)
+{
+ LogicalRepRelMapEntry *relentry;
+ LogicalRepTupleData *ridata;
+ ReplicaIdentityKey *rikey;
+ ReplicaIdentityEntry *rientry;
+ MemoryContext oldctx;
+ int n_ri;
+ bool found = false;
+
+ Assert(depends_on_xids);
+
+ /* Search for existing entry */
+ relentry = logicalrep_get_relentry(relid);
+
+ Assert(relentry);
+
+ /*
+ * First check whether any previous transaction (other than the current one)
+ * has affected the whole table e.g., truncate or schema change from
+ * publisher.
+ */
+ if (has_active_rel_dependency(relentry) &&
+ !TransactionIdEquals(relentry->last_depended_xid, new_depended_xid))
+ {
+ elog(DEBUG1, "found table-wide change affecting %u from %u",
+ relid, relentry->last_depended_xid);
+
+ append_xid_dependency(relentry->last_depended_xid, depends_on_xids);
+ }
+
+ n_ri = bms_num_members(relentry->remoterel.attkeys);
+
+ /*
+ * Return if there are no replica identity columns, indicating that the
+ * remote relation has neither a replica identity key nor is marked as
+ * replica identity full.
+ */
+ if (!n_ri)
+ return;
+
+ /*
+ * For replica identity FULL, the tuple contains all columns including
+ * NULLs, so we always check for existing dependencies.
+ *
+ * For other cases, NULL in the new tuple means the replica identity key
+ * hasn't changed, so no new dependency needs to be recorded. The dependency
+ * should have been recorded when processing the old tuple.
+ */
+ if (relentry->remoterel.replident != REPLICA_IDENTITY_FULL &&
+ has_null_key_values(original_data, relentry->remoterel.attkeys))
+ return;
+
+ oldctx = MemoryContextSwitchTo(ApplyContext);
+
+ /* Allocate space for replica identity values */
+ ridata = palloc0_object(LogicalRepTupleData);
+ ridata->colvalues = palloc0_array(StringInfoData, n_ri);
+ ridata->colstatus = palloc0_array(char, n_ri);
+ ridata->ncols = n_ri;
+
+ for (int i_original = 0, i_ri = 0; i_original < original_data->ncols; i_original++)
+ {
+ StringInfo original_colvalue = &original_data->colvalues[i_original];
+
+ if (!bms_is_member(i_original, relentry->remoterel.attkeys))
+ continue;
+
+ /*
+ * LOGICALREP_COLUMN_UNCHANGED only indicates that a TOAST column in the
+ * replica identity key hasn't changed. However, other columns may have
+ * changed, so we still need to check the dependency for this column.
+ *
+ * Before calling this function, unchanged TOAST column values should
+ * have been copied from the old tuple to the new tuple. So, we should
+ * see the complete replica identity key value in original_data and
+ * correctly check the dependency.
+ */
+ Assert(original_data->colstatus[i_original] != LOGICALREP_COLUMN_UNCHANGED ||
+ original_data->colvalues[i_original].len > 0);
+
+ if (ridata->colstatus[i_ri] != LOGICALREP_COLUMN_NULL)
+ {
+ initStringInfoExt(&ridata->colvalues[i_ri], original_colvalue->len + 1);
+ appendStringInfoString(&ridata->colvalues[i_ri], original_colvalue->data);
+ }
+
+ ridata->colstatus[i_ri] = original_data->colstatus[i_original];
+ i_ri++;
+ }
+
+ rikey = palloc0_object(ReplicaIdentityKey);
+ rikey->relid = relid;
+ rikey->data = ridata;
+
+ MemoryContextSwitchTo(oldctx);
+
+ /*
+ * The new xid could be invalid if the transaction will be applied by the
+ * leader itself which means all the changes will be committed before
+ * processing next transaction. In this case, we only need to check for
+ * dependencies on preceding transactions, there is no need to record a new
+ * dependency for subsequent transactions to wait on.
+ */
+ if (!TransactionIdIsValid(new_depended_xid))
+ {
+ rientry = replica_identity_lookup(replica_identity_table, rikey);
+ free_replica_identity_key(rikey);
+
+ if (rientry && has_active_key_dependency(rientry, true))
+ {
+ elog(DEBUG1, "found conflicting replica identity change on table %u from %u",
+ relid, rientry->remote_xid);
+
+ append_xid_dependency(rientry->remote_xid, depends_on_xids);
+ }
+
+ return;
+ }
+
+ /* Record a new dependency for subsequent transactions to wait on */
+ rientry = replica_identity_insert(replica_identity_table, rikey,
+ &found);
+
+ /*
+ * Release the key built to search the entry, if the entry already exists.
+ */
+ if (found)
+ {
+ free_replica_identity_key(rikey);
+
+ /*
+ * Append the dependency to the list if the current transaction was not
+ * the lastest one to modify the key.
+ */
+ if (has_active_key_dependency(rientry, false) &&
+ !TransactionIdEquals(rientry->remote_xid, new_depended_xid))
+ {
+ elog(DEBUG1, "found conflicting replica identity change on table %u from %u",
+ relid, rientry->remote_xid);
+
+ append_xid_dependency(rientry->remote_xid, depends_on_xids);
+ }
+ }
+
+ /* Update the new depended xid into the entry */
+ rientry->remote_xid = new_depended_xid;
+}
+
+/*
+ * Check for preceding transactions (other than 'current_xid') that involve
+ * insert, delete, or update operations on the specified table, and return them
+ * in 'depends_on_xids'.
+ */
+static void
+find_all_dependencies_on_rel(LogicalRepRelId relid,
+ TransactionId new_depended_xid,
+ List **depends_on_xids)
+{
+ replica_identity_iterator i;
+ ReplicaIdentityEntry *rientry;
+
+ Assert(depends_on_xids);
+
+ replica_identity_start_iterate(replica_identity_table, &i);
+ while ((rientry = replica_identity_iterate(replica_identity_table, &i)) != NULL)
+ {
+ Assert(TransactionIdIsValid(rientry->remote_xid));
+
+ if (rientry->keydata->relid != relid)
+ continue;
+
+ /* Skip entries that do not have a valid and uncommitted transaction */
+ if (!has_active_key_dependency(rientry, true))
+ continue;
+
+ /* Skip self-dependency */
+ if (TransactionIdEquals(rientry->remote_xid, new_depended_xid))
+ continue;
+
+ elog(DEBUG1, "found conflicting change on table %u from %u",
+ relid, rientry->remote_xid);
+
+ append_xid_dependency(rientry->remote_xid, depends_on_xids);
+ }
+}
+
+/*
+ * Check for any preceding transactions that affect the given table and returns
+ * them in 'depends_on_xids'.
+ *
+ * Additionally, if new_depended_xid is valid, record the current change and its
+ * transaction as a new table-level dependency, allowing subsequent transactions
+ * that modify the same table to be dependent on it.
+ */
+static void
+check_and_record_rel_dependency(LogicalRepRelId relid,
+ TransactionId new_depended_xid,
+ List **depends_on_xids)
+{
+ LogicalRepRelMapEntry *relentry;
+
+ Assert(depends_on_xids);
+
+ find_all_dependencies_on_rel(relid, new_depended_xid, depends_on_xids);
+
+ /* Search for existing entry */
+ relentry = logicalrep_get_relentry(relid);
+
+ /* First time seeing this relation, no entry exists. */
+ if (!relentry)
+ return;
+
+ /*
+ * Check whether any previous transaction (other than the current one) has
+ * affected the whole table e.g., truncate or schema change from publisher.
+ */
+ if (has_active_rel_dependency(relentry) &&
+ !TransactionIdEquals(relentry->last_depended_xid, new_depended_xid))
+ {
+ elog(DEBUG1, "found table-wide change affecting %u from %u",
+ relid, relentry->last_depended_xid);
+
+ append_xid_dependency(relentry->last_depended_xid, depends_on_xids);
+ }
+
+ if (TransactionIdIsValid(new_depended_xid))
+ relentry->last_depended_xid = new_depended_xid;
+}
+
+/*
+ * Check dependencies related to the current change by determining if the
+ * modification impacts the same row or table as another ongoing transaction.
+ *
+ * If a dependency on preceding transactions is found, this function instructs
+ * the appropriate worker (parallel apply or leader) to wait for those
+ * transactions to complete.
+ *
+ * Simultaneously, if the current change will be dispatched to a parallel apply
+ * worker (indicated by a valid new_depended_xid and a non-NULL winfo), it
+ * records a new dependency so that subsequent transactions will wait on it.
+ */
+static void
+handle_dependency_on_change(LogicalRepMsgType action, StringInfo s,
+ TransactionId new_depended_xid,
+ ParallelApplyWorkerInfo *winfo)
+{
+ LogicalRepRelId relid;
+ LogicalRepTupleData oldtup;
+ LogicalRepTupleData newtup;
+ LogicalRepRelation *rel;
+ List *depends_on_xids = NIL;
+ List *remote_relids;
+ bool has_oldtup = false;
+ bool cascade = false;
+ bool restart_seqs = false;
+
+ /*
+ * Parse the consume data using a local copy instead of directly consuming
+ * the given remote change as the caller may also read the data from the
+ * remote message.
+ */
+ StringInfoData change = *s;
+
+ /* Only the leader checks dependencies and schedules the parallel apply */
+ if (!am_leader_apply_worker())
+ return;
+
+ if (!ParallelApplyContext)
+ {
+ Assert(!replica_identity_table);
+
+ /*
+ * Create a permanent memory context for dependency information that
+ * persists across transactions. Using a dedicated context makes memory
+ * consumption more visible and easier to track, especially when
+ * handling a large number of change entries for transactions being
+ * applied in parallel.
+ */
+ ParallelApplyContext = AllocSetContextCreate(ApplyContext,
+ "ParallelApplyContext",
+ ALLOCSET_DEFAULT_SIZES);
+
+ replica_identity_table = replica_identity_create(ParallelApplyContext,
+ REPLICA_IDENTITY_INITIAL_SIZE,
+ NULL);
+ }
+
+ if (replica_identity_table->members >= REPLICA_IDENTITY_CLEANUP_THRESHOLD)
+ cleanup_committed_replica_identity_entries();
+
+ switch (action)
+ {
+ case LOGICAL_REP_MSG_INSERT:
+ relid = logicalrep_read_insert(&change, &newtup);
+ check_and_record_ri_dependency(relid, &newtup, new_depended_xid,
+ &depends_on_xids);
+ break;
+
+ case LOGICAL_REP_MSG_UPDATE:
+ relid = logicalrep_read_update(&change, &has_oldtup, &oldtup,
+ &newtup);
+
+ if (has_oldtup)
+ {
+ check_and_record_ri_dependency(relid, &oldtup, new_depended_xid,
+ &depends_on_xids);
+
+ /*
+ * Copy unchanged column values from the old tuple to the new
+ * tuple for replica identity dependency checking. See
+ * check_and_record_ri_dependency() for details. Also adjust the
+ * column status so that hash comparisons match correctly.
+ */
+ for (int i = 0; i < oldtup.ncols; i++)
+ {
+ if (newtup.colstatus[i] == LOGICALREP_COLUMN_UNCHANGED)
+ {
+ newtup.colvalues[i] = oldtup.colvalues[i];
+ newtup.colstatus[i] = oldtup.colstatus[i];
+ }
+ }
+ }
+
+ check_and_record_ri_dependency(relid, &newtup, new_depended_xid,
+ &depends_on_xids);
+ break;
+
+ case LOGICAL_REP_MSG_DELETE:
+ relid = logicalrep_read_delete(&change, &oldtup);
+ check_and_record_ri_dependency(relid, &oldtup, new_depended_xid,
+ &depends_on_xids);
+ break;
+
+ case LOGICAL_REP_MSG_TRUNCATE:
+ remote_relids = logicalrep_read_truncate(&change, &cascade,
+ &restart_seqs);
+
+ /*
+ * Truncate affects all rows in a table, so the current
+ * transaction should wait for all preceding transactions that
+ * modified the same table.
+ */
+ foreach_int(truncated_relid, remote_relids)
+ check_and_record_rel_dependency(truncated_relid,
+ new_depended_xid,
+ &depends_on_xids);
+
+ break;
+
+ case LOGICAL_REP_MSG_RELATION:
+ rel = logicalrep_read_rel(&change);
+
+ /*
+ * The replica identity key could be changed, making existing
+ * entries in the replica identity invalid. In this case, parallel
+ * apply is not allowed on this specific table until all running
+ * transactions that modified it have finished.
+ */
+ check_and_record_rel_dependency(rel->remoteid, new_depended_xid,
+ &depends_on_xids);
+ break;
+
+ case LOGICAL_REP_MSG_TYPE:
+ case LOGICAL_REP_MSG_MESSAGE:
+
+ /*
+ * Type updates accompany relation updates, so dependencies have
+ * already been checked during relation updates. Logical messages
+ * do not conflict with any changes, so they can be ignored.
+ */
+ break;
+
+ default:
+ Assert(false);
+ break;
+ }
+
+ /* Return early if the current change has no dependencies. */
+ if (!depends_on_xids)
+ return;
+
+ /*
+ * If the leader applies the transaction itself, start waiting for
+ * transactions that the current change depends on to finish. Otherwise,
+ * instruct the parallel apply worker to wait for them.
+ */
+ if (winfo == NULL)
+ {
+ foreach_xid(xid, depends_on_xids)
+ pa_wait_for_depended_transaction(xid);
+ }
+ else
+ {
+ send_internal_dependencies(winfo, depends_on_xids);
+ }
+}
+
/*
* Form the origin name for the subscription.
*
@@ -792,7 +1502,10 @@ handle_streamed_transaction(LogicalRepMsgType action, StringInfo s)
/* not in streaming mode */
if (apply_action == TRANS_LEADER_APPLY)
+ {
+ handle_dependency_on_change(action, s, InvalidTransactionId, winfo);
return false;
+ }
Assert(TransactionIdIsValid(stream_xid));
@@ -1244,6 +1957,29 @@ apply_handle_begin(StringInfo s)
pgstat_report_activity(STATE_RUNNING, NULL);
}
+/*
+ * Send an INTERNAL_DEPENDENCY message to a parallel apply worker.
+ */
+static void
+send_internal_dependencies(ParallelApplyWorkerInfo *winfo, List *depends_on_xids)
+{
+ StringInfoData dependencies;
+
+ initStringInfo(&dependencies);
+
+ pq_sendbyte(&dependencies, LOGICAL_REP_MSG_INTERNAL_MESSAGE);
+ pq_sendbyte(&dependencies, PA_MSG_XACT_DEPENDENCY);
+ pq_sendint32(&dependencies, list_length(depends_on_xids));
+
+ foreach_xid(xid, depends_on_xids)
+ pq_sendint32(&dependencies, xid);
+
+ if (pa_send_data(winfo, dependencies.len, dependencies.data))
+ ereport(ERROR,
+ errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
+ errmsg("could not send data to the logical replication parallel apply worker"));
+}
+
/*
* Handle COMMIT message.
*
@@ -1771,7 +2507,7 @@ apply_handle_stream_start(StringInfo s)
/* Try to allocate a worker for the streaming transaction. */
if (first_segment)
- pa_allocate_worker(stream_xid);
+ pa_allocate_worker(stream_xid, true);
apply_action = get_transaction_apply_action(stream_xid, &winfo);
@@ -3980,6 +4716,7 @@ store_flush_position(XLogRecPtr remote_lsn, XLogRecPtr local_lsn)
flushpos = palloc_object(FlushPosition);
flushpos->local_end = local_lsn;
flushpos->remote_end = remote_lsn;
+ flushpos->pa_remote_xid = InvalidTransactionId;
dlist_push_tail(&lsn_mapping, &flushpos->node);
MemoryContextSwitchTo(ApplyMessageContext);
diff --git a/src/include/replication/logicalrelation.h b/src/include/replication/logicalrelation.h
index efe0f9d6031..09890cef8d4 100644
--- a/src/include/replication/logicalrelation.h
+++ b/src/include/replication/logicalrelation.h
@@ -37,6 +37,14 @@ typedef struct LogicalRepRelMapEntry
/* Sync state. */
char state;
XLogRecPtr statelsn;
+
+ /*
+ * The last remote transaction that modified the relation's schema or
+ * truncated the relation. Used in dependency tracking to ensure subsequent
+ * transactions modifying the same table wait for this transaction to finish
+ * applying (see check_dependency_on_rel).
+ */
+ TransactionId last_depended_xid;
} LogicalRepRelMapEntry;
extern void logicalrep_relmap_update(LogicalRepRelation *remoterel);
@@ -50,5 +58,6 @@ extern void logicalrep_rel_close(LogicalRepRelMapEntry *rel,
LOCKMODE lockmode);
extern bool IsIndexUsableForReplicaIdentityFull(Relation idxrel, AttrMap *attrmap);
extern Oid GetRelationIdentityOrPK(Relation rel);
+extern LogicalRepRelMapEntry *logicalrep_get_relentry(LogicalRepRelId remoteid);
#endif /* LOGICALRELATION_H */
diff --git a/src/include/replication/worker_internal.h b/src/include/replication/worker_internal.h
index 43dcae43a62..3c88d9e1027 100644
--- a/src/include/replication/worker_internal.h
+++ b/src/include/replication/worker_internal.h
@@ -239,6 +239,8 @@ typedef struct ParallelApplyWorkerInfo
*/
bool in_use;
+ bool stream_txn;
+
ParallelApplyWorkerShared *shared;
} ParallelApplyWorkerInfo;
@@ -350,8 +352,9 @@ extern void apply_error_callback(void *arg);
extern void set_apply_error_context_origin(char *originname);
/* Parallel apply worker setup and interactions */
-extern void pa_allocate_worker(TransactionId xid);
+extern void pa_allocate_worker(TransactionId xid, bool stream_txn);
extern ParallelApplyWorkerInfo *pa_find_worker(TransactionId xid);
+extern XLogRecPtr pa_get_last_commit_end(TransactionId xid, bool *skipped_write);
extern void pa_detach_all_error_mq(void);
extern void apply_handle_internal_message(StringInfo s);
@@ -382,6 +385,7 @@ extern void pa_decr_and_wait_stream_block(void);
extern void pa_xact_finish(ParallelApplyWorkerInfo *winfo,
XLogRecPtr remote_lsn);
+extern bool pa_transaction_committed(TransactionId xid);
extern void pa_wait_for_depended_transaction(TransactionId xid);
diff --git a/src/test/subscription/meson.build b/src/test/subscription/meson.build
index e71e95c6297..fdc04d91f41 100644
--- a/src/test/subscription/meson.build
+++ b/src/test/subscription/meson.build
@@ -48,6 +48,7 @@ tests += {
't/036_sequences.pl',
't/037_except.pl',
't/038_walsnd_shutdown_timeout.pl',
+ 't/039_row_filter_toast_ri.pl',
't/100_bugs.pl',
],
},
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 0d5959aadb7..9609fd05967 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -2671,6 +2671,8 @@ ReplOriginXactState
ReplaceVarsFromTargetList_context
ReplaceVarsNoMatchOption
ReplaceWrapOption
+ReplicaIdentityEntry
+ReplicaIdentityKey
ReplicaIdentityStmt
ReplicationKind
ReplicationSlot
@@ -2682,6 +2684,7 @@ ReplicationSlotPersistentData
ReplicationState
ReplicationStateCtl
ReplicationStateOnDisk
+replica_identity_hash
ResTarget
ReservoirState
ReservoirStateData
--
2.43.0
[application/octet-stream] v21-0004-Parallel-apply-non-streaming-transactions.patch (67.9K, ../TY4PR01MB177180394D8E9C6D263E117E494FE2@TY4PR01MB17718.jpnprd01.prod.outlook.com/9-v21-0004-Parallel-apply-non-streaming-transactions.patch)
download | inline diff:
From 82efa9728f925d2bd4c8cb787996940a1bbec91f Mon Sep 17 00:00:00 2001
From: Zhijie Hou <[email protected]>
Date: Mon, 8 Jun 2026 10:16:46 +0800
Subject: [PATCH v21 4/9] Parallel apply non-streaming transactions
--
Basic design
--
The leader worker assigns each non-streaming transaction to a parallel apply
worker. Before dispatching changes to a parallel worker, the leader verifies if
the current modification affects the same row (identitied by replica identity
key) as another ongoing transaction. If so, the leader sends a list of dependent
transaction IDs to the parallel worker, indicating that the parallel apply
worker must wait for these transactions to commit before proceeding.
The leader preserves publisher commit order for all transactions by instructing
the parallel worker to wait for the last transaction to commit before committing
the current transaction. (Note that the leader itself does not wait the
parallelized transaction to be committed unless partial serialization mode is
active, where the parallel apply worker can no longer be reused.) See the
"commit order" and "worker interaction" sections below for details.
Tracking dependencies is necessary even when commit order is preserved.
Consider two transactions: TX-1 (INSERT row 1) and TX-2 (DELETE row 1). If
both are allowed to apply in parallel, TX-2's DELETE could be applied before
TX-1's INSERT, resulting in a delete_missing conflict.
Each parallel apply worker records the local end LSN of the transaction it
applies in shared memory. Subsequently, the leader gathers these local end LSNs
and logs them in the local 'lsn_mapping' for verifying whether they have been
flushed to disk (following the logic in get_flush_position()).
If no parallel apply worker is available, the leader will apply the transaction
independently.
For further details, please refer to the following:
--
dedendency tracking
--
The leader maintains a local hash table, using the remote change's replica
identity column values and relid as keys, with remote transaction IDs as values.
Before sending changes to the parallel apply worker, the leader computes a hash
using RI key values and the relid of the current change to search the hash
table. If an existing entry is found, the leader first updates the hash entry
with the receiving remote xid then tells the parallel worker to wait for it.
If the remote relation lacks a replica identity (RI), it indicates that only
INSERT can be replicated for this table. In such cases, the leader skips
dependency checks, allowing the parallel apply worker to proceed with applying
changes without delay. This is because the only potential conflict could happen
is related to the local unique key or foreign key, which that is yet to be
implemented.
In cases of TRUNCATE or remote schema changes affecting the entire table, the
leader retrieves all remote xids touching the same table (via sequential scans
of the hash table) and tells the parallel worker to wait for those transactions
to commit.
Hash entries are cleaned up once the transaction corresponding to the remote xid
in the entry has been committed. Clean-up typically occurs when collecting the
flush position of each transaction, but is forced if the hash table exceeds a
set threshold.
--
dedendency waiting
--
If a transaction is relied upon by others, the leader adds its xid to a shared
hash table. The shared hash table entry is cleared by the parallel apply worker
upon completing the transaction. Workers needing to wait for a transaction check
the shared hash table entry; if present, they lock the transaction ID (using
pa_lock_transaction). If absent, it indicates the transaction has been
committed, negating the need to wait.
--
commit order
--
We preserve publisher commit order for all transactions for two reasons:
1) User-visible consistency
Out-of-order commits can expose states on the subscriber that were never visible
on the publisher.
For example, suppose a user updates table A and then updates table B on the
publisher. If the subscriber commits those transactions out of order, a query
that sees the latest row in B might still see stale data in A. Although eventual
consistency would still be reached, that behavior may be unacceptable for some
users. In the future, we could provide a GUC to allow out-of-order commits for
users who prefer higher parallelism.
2) Replication progress tracking
We currently track replication progress using transaction commit LSNs. With
out-of-order commits, this becomes ambiguous after failures.
For example, if TX-2 is applied before TX-1 and replication stops due to an
error, we cannot reliably determine whether TX-1 was applied before restart.
As a result, transactions that were already committed on the subscriber may
be replayed.
--
worker interaction
--
After sending the COMMIT message for a transaction, the leader apply worker
does not wait for the parallel apply worker to finish applying that
transaction. Instead, it sends a PA_MSG_XACT_DEPENDENCY message to
the parallel apply worker, instructing it to wait for the last transaction to
commit. This allows the leader to remain busy receiving and dispatching
changes to more parallel apply workers, enabling greater parallelism in
transaction application.
Author: Zhijie Hou <[email protected]>
Author: Hayato Kuroda <[email protected]>
---
.../replication/logical/applyparallelworker.c | 348 +++++++++++++++--
src/backend/replication/logical/proto.c | 38 ++
src/backend/replication/logical/relation.c | 31 ++
src/backend/replication/logical/worker.c | 367 ++++++++++++++++--
src/include/replication/logicalproto.h | 2 +
src/include/replication/logicalrelation.h | 2 +
src/include/replication/worker_internal.h | 10 +-
src/test/subscription/meson.build | 2 +-
src/test/subscription/t/001_rep_changes.pl | 2 +
src/test/subscription/t/010_truncate.pl | 2 +-
src/test/subscription/t/015_stream.pl | 8 +-
src/test/subscription/t/026_stats.pl | 1 +
src/test/subscription/t/027_nosuperuser.pl | 1 +
src/test/subscription/t/050_parallel_apply.pl | 324 ++++++++++++++++
src/tools/pgindent/typedefs.list | 2 +
15 files changed, 1058 insertions(+), 82 deletions(-)
create mode 100644 src/test/subscription/t/050_parallel_apply.pl
diff --git a/src/backend/replication/logical/applyparallelworker.c b/src/backend/replication/logical/applyparallelworker.c
index 30d31e02ac1..4a7f65345d8 100644
--- a/src/backend/replication/logical/applyparallelworker.c
+++ b/src/backend/replication/logical/applyparallelworker.c
@@ -14,6 +14,9 @@
* ParallelApplyWorkerInfo which is required so the leader worker and parallel
* apply workers can communicate with each other.
*
+ * Streaming transactions
+ * ======================
+ *
* The parallel apply workers are assigned (if available) as soon as xact's
* first stream is received for subscriptions that have set their 'streaming'
* option as parallel. The leader apply worker will send changes to this new
@@ -36,12 +39,9 @@
* its information is added to the ParallelApplyWorkerPool. Once the worker
* finishes applying the transaction, it is marked as available for re-use.
* Now, before starting a new worker to apply the streaming transaction, we
- * check the list for any available worker. Note that we retain a maximum of
- * half the max_parallel_apply_workers_per_subscription workers in the pool and
- * after that, we simply exit the worker after applying the transaction.
- *
- * XXX This worker pool threshold is arbitrary and we can provide a GUC
- * variable for this in the future if required.
+ * check the list for any available worker. We do not stop workers in the pool
+ * unless partial of the changes of the transaction are serialized due to a send
+ * timeout.
*
* The leader apply worker will create a separate dynamic shared memory segment
* when each parallel apply worker starts. The reason for this design is that
@@ -152,6 +152,75 @@
* session-level locks because both locks could be acquired outside the
* transaction, and the stream lock in the leader needs to persist across
* transaction boundaries i.e. until the end of the streaming transaction.
+ *
+ * Non-streaming transactions
+ * ======================
+ * The handling is similar to streaming transactions, but including few
+ * differences:
+ *
+ * Transaction dependency
+ * ----------------------
+ * Before dispatching changes to a parallel worker, the leader verifies if the
+ * current modification affects the same row (identitied by replica identity
+ * key) as another ongoing transaction (see handle_dependency_on_change for
+ * details). If so, the leader sends a list of dependent transaction IDs to the
+ * parallel worker, indicating that the parallel apply worker must wait for
+ * these transactions to commit before proceeding.
+ *
+ * Tracking dependencies is necessary even when commit order is preserved.
+ * Consider two transactions: TX-1 (INSERT row 1) and TX-2 (DELETE row 1). If
+ * both are allowed to apply in parallel, TX-2's DELETE could be applied before
+ * TX-1's INSERT, resulting in a delete_missing conflict.
+ *
+ * Commit order
+ * ------------
+ * We preserve publisher commit order for all transactions for two reasons:
+ *
+ * 1) User-visible consistency
+ *
+ * Out-of-order commits can expose states on the subscriber that were never
+ * visible on the publisher.
+ *
+ * For example, suppose a user updates table A and then updates table B on the
+ * publisher. If the subscriber commits those transactions out of order, a
+ * query that sees the latest row in B might still see stale data in A.
+ * Although eventual consistency would still be reached, that behavior may be
+ * unacceptable for some users. In the future, we could provide a subscription
+ * option to allow out-of-order commits for users who prefer higher parallelism.
+ *
+ * 2) Replication progress tracking
+ *
+ * We currently track replication progress using the last transaction's commit
+ * LSNs. With out-of-order commits, this becomes ambiguous after failures.
+ *
+ * For example, if TX-2 is applied before TX-1 and replication stops due to an
+ * error, we cannot reliably determine whether TX-1 was applied before restart.
+ * As a result, transactions that were already committed on the subscriber may
+ * be replayed.
+ *
+ * Worker interaction
+ * ------------
+ * After sending the COMMIT message for a transaction, the leader apply worker
+ * does not wait for the parallel apply worker to finish applying that
+ * transaction. Instead, it sends a PA_MSG_XACT_DEPENDENCY message to
+ * the parallel apply worker, instructing it to wait for the last transaction to
+ * commit. This allows the leader to remain busy receiving and dispatching
+ * changes to more parallel apply workers, enabling greater parallelism in
+ * transaction application.
+ *
+ * Locking considerations
+ * ----------------------
+ * When handling a PA_MSG_RELMAP message, the worker attempts
+ * to acquire the transaction lock of the depended transaction and releases it
+ * immediately after acquisition (see pa_wait_for_depended_transaction). This
+ * allows deadlock detection when one worker (either leader or parallel apply
+ * worker) is waiting for a dependency on a transaction being applied by another
+ * worker, while that other worker is also blocked by a lock held by the first
+ * worker.
+ *
+ * The lock graph for the above example will look as follows: Worker_1 (waiting
+ * for depended transaction to finish) -> Worker_2 (waiting to acquire a
+ * relation lock) -> Worker_1
*-------------------------------------------------------------------------
*/
@@ -316,6 +385,7 @@ static ParallelTransState pa_get_xact_state(ParallelApplyWorkerShared *wshared);
static PartialFileSetState pa_get_fileset_state(void);
static void pa_attach_parallelized_txn_hash(dsa_handle *pa_dsa_handle,
dshash_table_handle *pa_dshash_handle);
+static void write_internal_relation(StringInfo s, LogicalRepRelation *rel);
/*
* Returns true if it is OK to start a parallel apply worker, false otherwise.
@@ -433,6 +503,7 @@ pa_setup_dsm(ParallelApplyWorkerInfo *winfo)
shared = shm_toc_allocate(toc, sizeof(ParallelApplyWorkerShared));
SpinLockInit(&shared->mutex);
+ shared->xid = InvalidTransactionId;
shared->xact_state = PARALLEL_TRANS_UNKNOWN;
pg_atomic_init_u32(&(shared->pending_stream_count), 0);
shared->last_commit_end = InvalidXLogRecPtr;
@@ -476,6 +547,8 @@ pa_launch_parallel_worker(void)
MemoryContext oldcontext;
bool launched;
ParallelApplyWorkerInfo *winfo;
+ dsa_handle pa_dsa_handle;
+ dshash_table_handle pa_dshash_handle;
ListCell *lc;
/* Try to get an available parallel apply worker from the worker pool. */
@@ -485,8 +558,36 @@ pa_launch_parallel_worker(void)
if (!winfo->in_use)
return winfo;
+
+ /*
+ * The leader does not explicitly free workers that handle non-streaming
+ * transactions, so we check whether the transaction has committed, and
+ * if so, reuse the worker.
+ */
+ if (!winfo->stream_txn &&
+ pa_get_xact_state(winfo->shared) == PARALLEL_TRANS_FINISHED)
+ {
+ /*
+ * Save the local commit LSN of the last transaction applied by
+ * this worker before reusing it for another transaction. This WAL
+ * position is crucial for determining the flush position in
+ * responses to the publisher (see get_flush_position()).
+ */
+ (void) pa_get_last_commit_end(winfo->shared->xid, false, NULL);
+ return winfo;
+ }
}
+ pa_attach_parallelized_txn_hash(&pa_dsa_handle, &pa_dshash_handle);
+
+ /*
+ * Return if the number of parallel apply workers has reached the maximum
+ * limit.
+ */
+ if (list_length(ParallelApplyWorkerPool) ==
+ max_parallel_apply_workers_per_subscription)
+ return NULL;
+
/*
* Start a new parallel apply worker.
*
@@ -514,18 +615,50 @@ pa_launch_parallel_worker(void)
dsm_segment_handle(winfo->dsm_seg),
false);
- if (launched)
- {
- ParallelApplyWorkerPool = lappend(ParallelApplyWorkerPool, winfo);
- }
- else
+ if (!launched)
{
+ MemoryContextSwitchTo(oldcontext);
pa_free_worker_info(winfo);
- winfo = NULL;
+ return NULL;
}
+ ParallelApplyWorkerPool = lappend(ParallelApplyWorkerPool, winfo);
+
MemoryContextSwitchTo(oldcontext);
+ /*
+ * Send all existing remote relation information to the parallel apply
+ * worker. This allows the parallel worker to initialize the
+ * LogicalRepRelMapEntry locally before applying remote changes. This is
+ * needed since the walsender does not send remote relation information
+ * with every transaction.
+ */
+ if (logicalrep_get_num_rels())
+ {
+ StringInfoData out;
+ shm_mq_result result;
+
+ initStringInfo(&out);
+
+ write_internal_relation(&out, NULL);
+
+ /*
+ * Send relation information synchronously rather than using
+ * non-blocking send (pa_send_data). This avoids the need to serialize
+ * the data, which would add complexity for the parallel apply worker to
+ * deserialize outside of transaction application.
+ *
+ * This is safe from deadlocks because a newly started worker does not
+ * hold any strong locks when processing relation information.
+ */
+ result = shm_mq_send(winfo->mq_handle, out.len, out.data, false, true);
+
+ if (result != SHM_MQ_SUCCESS)
+ ereport(ERROR,
+ errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
+ errmsg("could not send remote relation information to the logical replication parallel apply worker"));
+ }
+
return winfo;
}
@@ -630,25 +763,27 @@ pa_free_worker(ParallelApplyWorkerInfo *winfo)
{
Assert(!am_parallel_apply_worker());
Assert(winfo->in_use);
- Assert(pa_get_xact_state(winfo->shared) == PARALLEL_TRANS_FINISHED);
+ Assert(!winfo->stream_txn ||
+ pa_get_xact_state(winfo->shared) == PARALLEL_TRANS_FINISHED);
if (!hash_search(ParallelApplyTxnHash, &winfo->shared->xid, HASH_REMOVE, NULL))
elog(ERROR, "hash table corrupted");
/*
- * Stop the worker if there are enough workers in the pool.
+ * XXX we stop the worker if the leader apply worker serialize part of the
+ * transaction data due to a send timeout. This is because the message could
+ * be partially written to the queue and there is no way to clean the queue
+ * other than resending the message until it succeeds. Instead of trying to
+ * send the data which anyway would have been serialized and then letting
+ * the parallel apply worker deal with the spurious message, we stop the
+ * worker.
*
- * XXX Additionally, we also stop the worker if the leader apply worker
- * serialize part of the transaction data due to a send timeout. This is
- * because the message could be partially written to the queue and there
- * is no way to clean the queue other than resending the message until it
- * succeeds. Instead of trying to send the data which anyway would have
- * been serialized and then letting the parallel apply worker deal with
- * the spurious message, we stop the worker.
+ * For other cases, we do not stop workers once started. All transactions
+ * (whether streamed or not) are assigned to parallel apply workers, so
+ * restarting workers frequently would only increase CPU overhead and slow
+ * down the leader's ability to dispatch changes to workers.
*/
- if (winfo->serialize_changes ||
- list_length(ParallelApplyWorkerPool) >
- (max_parallel_apply_workers_per_subscription / 2))
+ if (winfo->serialize_changes)
{
logicalrep_pa_worker_stop(winfo);
pa_free_worker_info(winfo);
@@ -781,15 +916,17 @@ pa_process_spooled_messages_if_required(void)
/*
* Get the local end LSN for a transaction applied by a parallel apply worker.
*
+ * Set delete_entry to true if you intend to remove the transaction from the
+ * ParallelApplyTxnHash after collecting its LSN. Otherwise, the local end LSN
+ * is copied from shared memory into the local entry of ParallelApplyTxnHash for
+ * later collection by the leader.
+ *
* If the parallel apply worker did not write any changes during the transaction
* application due to situations like update/delete_missing or a before trigger,
* the *skipped_write will be set to true.
- *
- * Once the LSN is retrieved, the transaction entry is removed from
- * ParallelApplyTxnHash.
*/
XLogRecPtr
-pa_get_last_commit_end(TransactionId xid, bool *skipped_write)
+pa_get_last_commit_end(TransactionId xid, bool delete_entry, bool *skipped_write)
{
bool found;
ParallelApplyWorkerEntry *entry;
@@ -815,7 +952,8 @@ pa_get_last_commit_end(TransactionId xid, bool *skipped_write)
winfo = entry->winfo;
if (winfo == NULL)
{
- if (!hash_search(ParallelApplyTxnHash, &xid, HASH_REMOVE, NULL))
+ if (delete_entry &&
+ !hash_search(ParallelApplyTxnHash, &xid, HASH_REMOVE, NULL))
elog(ERROR, "hash table corrupted");
if (skipped_write)
@@ -838,7 +976,8 @@ pa_get_last_commit_end(TransactionId xid, bool *skipped_write)
elog(DEBUG1, "store local commit %X/%X end to txn entry: %u",
LSN_FORMAT_ARGS(entry->local_end), xid);
- if (!hash_search(ParallelApplyTxnHash, &xid, HASH_REMOVE, NULL))
+ if (delete_entry &&
+ !hash_search(ParallelApplyTxnHash, &xid, HASH_REMOVE, NULL))
elog(ERROR, "hash table corrupted");
return entry->local_end;
@@ -1091,6 +1230,9 @@ pa_shutdown(int code, Datum arg)
INVALID_PROC_NUMBER);
dsm_detach((dsm_segment *) DatumGetPointer(arg));
+
+ if (parallel_apply_dsa_area)
+ dsa_detach(parallel_apply_dsa_area);
}
/*
@@ -1402,7 +1544,6 @@ pa_send_data(ParallelApplyWorkerInfo *winfo, Size nbytes, const void *data)
shm_mq_result result;
TimestampTz startTime = 0;
- Assert(!IsTransactionState());
Assert(!winfo->serialize_changes);
/*
@@ -1454,6 +1595,58 @@ pa_send_data(ParallelApplyWorkerInfo *winfo, Size nbytes, const void *data)
}
}
+/*
+ * Distribute remote relation information to all active parallel apply workers
+ * that require it.
+ */
+void
+pa_distribute_schema_changes_to_workers(LogicalRepRelation *rel)
+{
+ StringInfoData out;
+
+ /* Only leader apply workers can distribute schema changes */
+ if (!am_leader_apply_worker())
+ return;
+
+ /* Quick exit if there are no parallel apply workers */
+ if (!ParallelApplyWorkerPool)
+ return;
+
+ initStringInfo(&out);
+
+ write_internal_relation(&out, rel);
+
+ foreach_ptr(ParallelApplyWorkerInfo, winfo, ParallelApplyWorkerPool)
+ {
+ /*
+ * Skip the worker responsible for the current transaction, as the
+ * relation information has already been sent to it.
+ */
+ if (winfo == stream_apply_worker)
+ continue;
+
+ /*
+ * Skip the worker that is in serialize mode, as they will soon stop
+ * once they finish applying the transaction.
+ */
+ if (winfo->serialize_changes)
+ continue;
+
+ elog(DEBUG1, "distributing schema changes to pa workers");
+
+ if (pa_send_data(winfo, out.len, out.data))
+ continue;
+
+ /*
+ * TODO: Support switching to PARTIAL_SERIALIZE mode when the send
+ * buffer becomes full.
+ */
+ ereport(ERROR,
+ errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
+ errmsg("could not send data to the logical replication parallel apply worker for subscription"));
+ }
+}
+
/*
* Switch to PARTIAL_SERIALIZE mode for the current transaction -- this means
* that the current data and any subsequent data for this transaction will be
@@ -1870,25 +2063,41 @@ pa_decr_and_wait_stream_block(void)
void
pa_xact_finish(ParallelApplyWorkerInfo *winfo, XLogRecPtr remote_lsn)
{
+ XLogRecPtr local_lsn = InvalidXLogRecPtr;
+ TransactionId pa_remote_xid = winfo->shared->xid;
+
Assert(am_leader_apply_worker());
/*
- * Unlock the shared object lock so that parallel apply worker can
- * continue to receive and apply changes.
+ * Unlock the shared object lock taken for streaming transactions so that
+ * parallel apply worker can continue to receive and apply changes.
*/
- pa_unlock_stream(winfo->shared->xid, AccessExclusiveLock);
+ if (winfo->stream_txn)
+ pa_unlock_stream(winfo->shared->xid, AccessExclusiveLock);
/*
- * Wait for that worker to finish. This is necessary to maintain commit
- * order which avoids failures due to transaction dependencies and
- * deadlocks.
+ * Wait for that worker for streaming transaction to finish. This is
+ * necessary to maintain commit order which avoids failures due to
+ * transaction dependencies and deadlocks.
+ *
+ * For transactions in partial seralize mode, wait for stop as well as the
+ * worker is anyway cannot be reused anymore (see pa_free_worker() for
+ * details).
*/
- pa_wait_for_xact_finish(winfo);
+ if (winfo->serialize_changes || winfo->stream_txn)
+ {
+ pa_wait_for_xact_finish(winfo);
+
+ local_lsn = winfo->shared->last_commit_end;
+ pa_remote_xid = InvalidTransactionId;
+
+ pa_free_worker(winfo);
+ }
if (XLogRecPtrIsValid(remote_lsn))
- store_flush_position(remote_lsn, winfo->shared->last_commit_end);
+ store_flush_position(remote_lsn, local_lsn, pa_remote_xid);
- pa_free_worker(winfo);
+ pa_set_stream_apply_worker(NULL);
}
/*
@@ -2022,3 +2231,62 @@ pa_wait_for_depended_transaction(TransactionId xid)
elog(DEBUG1, "finish waiting for depended xid %u", xid);
}
+
+/*
+ * Mark the transaction state as finished and remove the shared hash entry.
+ */
+void
+pa_commit_transaction(void)
+{
+ TransactionId xid = MyParallelShared->xid;
+
+ SpinLockAcquire(&MyParallelShared->mutex);
+ MyParallelShared->xact_state = PARALLEL_TRANS_FINISHED;
+ SpinLockRelease(&MyParallelShared->mutex);
+
+ dshash_delete_key(parallelized_txns, &xid);
+ elog(DEBUG1, "xid %u committed", xid);
+}
+
+/*
+ * Write internal relation description to the output stream.
+ */
+static void
+write_internal_relation(StringInfo s, LogicalRepRelation *rel)
+{
+ pq_sendbyte(s, LOGICAL_REP_MSG_INTERNAL_MESSAGE);
+ pq_sendbyte(s, PA_MSG_RELMAP);
+
+ if (rel)
+ {
+ pq_sendint(s, 1, 4);
+ logicalrep_write_internal_rel(s, rel);
+ }
+ else
+ {
+ pq_sendint(s, logicalrep_get_num_rels(), 4);
+ logicalrep_write_all_rels(s);
+ }
+}
+
+/*
+ * Register a transaction to the shared hash table.
+ *
+ * This function is called by the leader during the commit phase of non-streamed
+ * transactions. The parallel apply worker that applies the transaction will
+ * remove it from the hash table upon completion.
+ */
+void
+pa_add_parallelized_transaction(TransactionId xid)
+{
+ bool found;
+ ParallelizedTxnEntry *txn_entry;
+
+ Assert(parallelized_txns);
+ Assert(TransactionIdIsValid(xid));
+ Assert(am_leader_apply_worker());
+
+ txn_entry = dshash_find_or_insert(parallelized_txns, &xid, &found);
+
+ dshash_release_lock(parallelized_txns, txn_entry);
+}
diff --git a/src/backend/replication/logical/proto.c b/src/backend/replication/logical/proto.c
index a0cac05cdcd..eb1a846ce0c 100644
--- a/src/backend/replication/logical/proto.c
+++ b/src/backend/replication/logical/proto.c
@@ -691,6 +691,44 @@ logicalrep_write_rel(StringInfo out, TransactionId xid, Relation rel,
logicalrep_write_attrs(out, rel, columns, include_gencols_type);
}
+/*
+ * Write internal relation description to the output stream.
+ */
+void
+logicalrep_write_internal_rel(StringInfo out, LogicalRepRelation *rel)
+{
+ pq_sendint32(out, rel->remoteid);
+
+ /* Write relation name */
+ pq_sendstring(out, rel->nspname);
+ pq_sendstring(out, rel->relname);
+
+ /* Write the replica identity. */
+ pq_sendbyte(out, rel->replident);
+
+ /* Write attribute description */
+ pq_sendint16(out, rel->natts);
+
+ for (int i = 0; i < rel->natts; i++)
+ {
+ uint8 flags = 0;
+
+ if (bms_is_member(i, rel->attkeys))
+ flags |= LOGICALREP_IS_REPLICA_IDENTITY;
+
+ pq_sendbyte(out, flags);
+
+ /* attribute name */
+ pq_sendstring(out, rel->attnames[i]);
+
+ /* attribute type id */
+ pq_sendint32(out, rel->atttyps[i]);
+
+ /* ignore attribute mode for now */
+ pq_sendint32(out, 0);
+ }
+}
+
/*
* Read the relation info from stream and return as LogicalRepRelation.
*/
diff --git a/src/backend/replication/logical/relation.c b/src/backend/replication/logical/relation.c
index 774e048c263..c5a83b095fa 100644
--- a/src/backend/replication/logical/relation.c
+++ b/src/backend/replication/logical/relation.c
@@ -960,6 +960,37 @@ FindLogicalRepLocalIndex(Relation localrel, LogicalRepRelation *remoterel,
return InvalidOid;
}
+/*
+ * Get the number of entries in the LogicalRepRelMap.
+ */
+int
+logicalrep_get_num_rels(void)
+{
+ if (LogicalRepRelMap == NULL)
+ return 0;
+
+ return hash_get_num_entries(LogicalRepRelMap);
+}
+
+/*
+ * Write all the remote relation information from the LogicalRepRelMapEntry to
+ * the output stream.
+ */
+void
+logicalrep_write_all_rels(StringInfo out)
+{
+ LogicalRepRelMapEntry *entry;
+ HASH_SEQ_STATUS status;
+
+ if (LogicalRepRelMap == NULL)
+ return;
+
+ hash_seq_init(&status, LogicalRepRelMap);
+
+ while ((entry = (LogicalRepRelMapEntry *) hash_seq_search(&status)) != NULL)
+ logicalrep_write_internal_rel(out, &entry->remoterel);
+}
+
/*
* Get the LogicalRepRelMapEntry corresponding to the given relid without
* opening the local relation.
diff --git a/src/backend/replication/logical/worker.c b/src/backend/replication/logical/worker.c
index 3a1423fbe58..b5013cbfc00 100644
--- a/src/backend/replication/logical/worker.c
+++ b/src/backend/replication/logical/worker.c
@@ -290,6 +290,7 @@
#include "tcop/tcopprot.h"
#include "utils/acl.h"
#include "utils/guc.h"
+#include "utils/injection_point.h"
#include "utils/inval.h"
#include "utils/lsyscache.h"
#include "utils/memutils.h"
@@ -514,6 +515,8 @@ static List *on_commit_wakeup_workers_subids = NIL;
bool in_remote_transaction = false;
static XLogRecPtr remote_final_lsn = InvalidXLogRecPtr;
+static TransactionId remote_xid = InvalidTransactionId;
+static TransactionId last_parallelized_remote_xid = InvalidTransactionId;
/* fields valid only when processing streamed transaction */
static bool in_streamed_transaction = false;
@@ -723,6 +726,7 @@ static void on_exit_clear_xact_state(int code, Datum arg);
static void send_internal_dependencies(ParallelApplyWorkerInfo *winfo,
List *depends_on_xids);
+static void build_dependency_with_last_committed_txn(ParallelApplyWorkerInfo *winfo);
/*
* Compute the hash value for entries in the replica_identity_table.
@@ -765,6 +769,9 @@ hash_replica_identity_compare(ReplicaIdentityKey *a, ReplicaIdentityKey *b)
if (a->data->colstatus[i] != b->data->colstatus[i])
return false;
+ if (a->data->colstatus[i] == LOGICALREP_COLUMN_NULL)
+ continue;
+
if (a->data->colvalues[i].len != b->data->colvalues[i].len)
return false;
@@ -842,7 +849,7 @@ cleanup_committed_replica_identity_entries(void)
XLogRecPtrIsValid(pos->local_end))
continue;
- pos->local_end = pa_get_last_commit_end(pos->pa_remote_xid,
+ pos->local_end = pa_get_last_commit_end(pos->pa_remote_xid, false,
&skipped_write);
elog(DEBUG1,
@@ -1011,8 +1018,6 @@ check_and_record_ri_dependency(Oid relid, LogicalRepTupleData *original_data,
for (int i_original = 0, i_ri = 0; i_original < original_data->ncols; i_original++)
{
- StringInfo original_colvalue = &original_data->colvalues[i_original];
-
if (!bms_is_member(i_original, relentry->remoterel.attkeys))
continue;
@@ -1029,8 +1034,10 @@ check_and_record_ri_dependency(Oid relid, LogicalRepTupleData *original_data,
Assert(original_data->colstatus[i_original] != LOGICALREP_COLUMN_UNCHANGED ||
original_data->colvalues[i_original].len > 0);
- if (ridata->colstatus[i_ri] != LOGICALREP_COLUMN_NULL)
+ if (original_data->colstatus[i_original] != LOGICALREP_COLUMN_NULL)
{
+ StringInfo original_colvalue = &original_data->colvalues[i_original];
+
initStringInfoExt(&ridata->colvalues[i_ri], original_colvalue->len + 1);
appendStringInfoString(&ridata->colvalues[i_ri], original_colvalue->data);
}
@@ -1498,14 +1505,14 @@ handle_streamed_transaction(LogicalRepMsgType action, StringInfo s)
TransApplyAction apply_action;
StringInfoData original_msg;
+ if (!in_streamed_transaction)
+ return false;
+
apply_action = get_transaction_apply_action(stream_xid, &winfo);
/* not in streaming mode */
if (apply_action == TRANS_LEADER_APPLY)
- {
- handle_dependency_on_change(action, s, InvalidTransactionId, winfo);
return false;
- }
Assert(TransactionIdIsValid(stream_xid));
@@ -1580,6 +1587,73 @@ handle_streamed_transaction(LogicalRepMsgType action, StringInfo s)
}
}
+/*
+ * Handle non-streaming transactions when parallel apply is in use.
+ *
+ * This function runs only in the leader apply worker while processing a remote
+ * transaction. It checks whether the current change has dependencies on
+ * preceding parallelized transactions and decides whether to send the change to
+ * a parallel apply worker.
+ *
+ * Returns true if the change has been dispatched to a parallel worker,
+ * indicating the leader does not need to apply it directly. Returns false
+ * otherwise.
+ *
+ * Exception: If the message being processed is LOGICAL_REP_MSG_RELATION or
+ * LOGICAL_REP_MSG_TYPE, return false even if the message needs to be sent to a
+ * parallel apply worker.
+ */
+static bool
+handle_parallelized_transaction(LogicalRepMsgType action, StringInfo s)
+{
+ ParallelApplyWorkerInfo *winfo;
+ TransApplyAction apply_action;
+
+ /*
+ * Dependency checking for non-streaming transactions is only required in
+ * the leader apply worker during a remote transaction.
+ */
+ if (!in_remote_transaction || !am_leader_apply_worker())
+ return false;
+
+ apply_action = get_transaction_apply_action(remote_xid, &winfo);
+
+ /* not assigned to parallel apply worker, apply in leader */
+ if (apply_action == TRANS_LEADER_APPLY)
+ {
+ handle_dependency_on_change(action, s, InvalidTransactionId, winfo);
+ return false;
+ }
+
+ Assert(TransactionIdIsValid(remote_xid));
+
+ handle_dependency_on_change(action, s, remote_xid, winfo);
+
+ switch (apply_action)
+ {
+ case TRANS_LEADER_SEND_TO_PARALLEL:
+ Assert(winfo);
+
+ /* Always update relation and type cache in leader apply worker */
+ if (pa_send_data(winfo, s->len, s->data))
+ return (action != LOGICAL_REP_MSG_RELATION &&
+ action != LOGICAL_REP_MSG_TYPE);
+
+ /*
+ * TODO: Support switching to PARTIAL_SERIALIZE mode when the send
+ * buffer becomes full.
+ */
+ ereport(ERROR,
+ errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
+ errmsg("could not send data to the logical replication parallel apply worker"));
+ return false; /* silence compiler warning */
+
+ default:
+ elog(ERROR, "unexpected apply action: %d", (int) apply_action);
+ return false; /* silence compiler warning */
+ }
+}
+
/*
* Executor state preparation for evaluation of constraint expressions,
* indexes and triggers for the specified relation.
@@ -1941,17 +2015,61 @@ static void
apply_handle_begin(StringInfo s)
{
LogicalRepBeginData begin_data;
+ ParallelApplyWorkerInfo *winfo;
+ TransApplyAction apply_action;
/* There must not be an active streaming transaction. */
Assert(!TransactionIdIsValid(stream_xid));
logicalrep_read_begin(s, &begin_data);
- set_apply_error_context_xact(begin_data.xid, begin_data.final_lsn);
+
+ remote_xid = begin_data.xid;
+
+ set_apply_error_context_xact(remote_xid, begin_data.final_lsn);
remote_final_lsn = begin_data.final_lsn;
maybe_start_skipping_changes(begin_data.final_lsn);
+ pa_allocate_worker(remote_xid, false);
+
+ apply_action = get_transaction_apply_action(remote_xid, &winfo);
+
+ elog(DEBUG1, "new remote_xid %u", remote_xid);
+ switch (apply_action)
+ {
+ case TRANS_LEADER_APPLY:
+ break;
+
+ case TRANS_LEADER_SEND_TO_PARALLEL:
+ Assert(winfo);
+
+ if (pa_send_data(winfo, s->len, s->data))
+ {
+ pa_set_stream_apply_worker(winfo);
+ break;
+ }
+
+ /*
+ * TODO: Support switching to PARTIAL_SERIALIZE mode when the send
+ * buffer becomes full.
+ */
+ ereport(ERROR,
+ errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
+ errmsg("could not send data to the logical replication parallel apply worker"));
+ break;
+
+ case TRANS_PARALLEL_APPLY:
+ /* Hold the lock until the end of the transaction. */
+ pa_lock_transaction(MyParallelShared->xid, AccessExclusiveLock);
+ pa_set_xact_state(MyParallelShared, PARALLEL_TRANS_STARTED);
+ break;
+
+ default:
+ elog(ERROR, "unexpected apply action: %d", (int) apply_action);
+ break;
+ }
+
in_remote_transaction = true;
pgstat_report_activity(STATE_RUNNING, NULL);
@@ -1974,12 +2092,33 @@ send_internal_dependencies(ParallelApplyWorkerInfo *winfo, List *depends_on_xids
foreach_xid(xid, depends_on_xids)
pq_sendint32(&dependencies, xid);
- if (pa_send_data(winfo, dependencies.len, dependencies.data))
+ /*
+ * TODO: Support switching to PARTIAL_SERIALIZE mode when the send
+ * buffer becomes full.
+ */
+ if (!pa_send_data(winfo, dependencies.len, dependencies.data))
ereport(ERROR,
errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
errmsg("could not send data to the logical replication parallel apply worker"));
}
+/*
+ * Make a dependency between this and the lastly committed transaction.
+ *
+ * Sends an INTERNAL_DEPENDENCY message to the parallel apply worker,
+ * instructing it to wait for the last committed transaction to finish before
+ * committing its own, thereby preserving commit order.
+ */
+static void
+build_dependency_with_last_committed_txn(ParallelApplyWorkerInfo *winfo)
+{
+ /* Skip if transactions have not been applied yet */
+ if (!TransactionIdIsValid(last_parallelized_remote_xid))
+ return;
+
+ send_internal_dependencies(winfo, list_make1_xid(last_parallelized_remote_xid));
+}
+
/*
* Handle COMMIT message.
*
@@ -1989,6 +2128,8 @@ static void
apply_handle_commit(StringInfo s)
{
LogicalRepCommitData commit_data;
+ ParallelApplyWorkerInfo *winfo;
+ TransApplyAction apply_action;
logicalrep_read_commit(s, &commit_data);
@@ -1999,7 +2140,92 @@ apply_handle_commit(StringInfo s)
LSN_FORMAT_ARGS(commit_data.commit_lsn),
LSN_FORMAT_ARGS(remote_final_lsn))));
- apply_handle_commit_internal(&commit_data);
+ apply_action = get_transaction_apply_action(remote_xid, &winfo);
+
+ switch (apply_action)
+ {
+ case TRANS_LEADER_APPLY:
+
+ /*
+ * Apart from parallelized transactions, we do not have to
+ * register this transaction to parallelized_txns. The commit
+ * ordering is always preserved.
+ */
+
+ /* Wait until the last transaction finishes */
+ if (TransactionIdIsValid(last_parallelized_remote_xid))
+ {
+ pa_wait_for_depended_transaction(last_parallelized_remote_xid);
+ last_parallelized_remote_xid = InvalidTransactionId;
+ }
+
+ apply_handle_commit_internal(&commit_data);
+
+ break;
+
+ case TRANS_LEADER_SEND_TO_PARALLEL:
+ Assert(winfo);
+
+ /*
+ * Mark this transaction as parallelized. This ensures that
+ * upcoming transactions wait until this transaction is committed.
+ */
+ pa_add_parallelized_transaction(remote_xid);
+
+ /*
+ * Build a dependency between this transaction and the lastly
+ * committed transaction to preserve the commit order.
+ */
+ build_dependency_with_last_committed_txn(winfo);
+
+ if (pa_send_data(winfo, s->len, s->data))
+ {
+ /* Cache the remote_xid */
+ last_parallelized_remote_xid = remote_xid;
+
+ /* Finish processing the transaction. */
+ pa_xact_finish(winfo, commit_data.end_lsn);
+ break;
+ }
+
+ /*
+ * TODO: Support switching to PARTIAL_SERIALIZE mode when the send
+ * buffer becomes full.
+ */
+ ereport(ERROR,
+ errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
+ errmsg("could not send data to the logical replication parallel apply worker"));
+ break;
+
+ case TRANS_PARALLEL_APPLY:
+
+ /*
+ * If the parallel apply worker is applying spooled messages then
+ * close the file before committing.
+ */
+ if (stream_fd)
+ stream_close_file();
+
+ INJECTION_POINT("parallel-worker-before-commit", NULL);
+
+ apply_handle_commit_internal(&commit_data);
+
+ MyParallelShared->last_commit_end = XactLastCommitEnd;
+
+ pa_commit_transaction();
+
+ pa_unlock_transaction(remote_xid, AccessExclusiveLock);
+ break;
+
+ default:
+ elog(ERROR, "unexpected apply action: %d", (int) apply_action);
+ break;
+ }
+
+ remote_xid = InvalidTransactionId;
+ in_remote_transaction = false;
+
+ elog(DEBUG1, "reset remote_xid %u", remote_xid);
/*
* Process any tables that are being synchronized in parallel, as well as
@@ -2122,7 +2348,8 @@ apply_handle_prepare(StringInfo s)
* XactLastCommitEnd, and adding it for this purpose doesn't seems worth
* it.
*/
- store_flush_position(prepare_data.end_lsn, InvalidXLogRecPtr);
+ store_flush_position(prepare_data.end_lsn, InvalidXLogRecPtr,
+ InvalidTransactionId);
in_remote_transaction = false;
@@ -2182,7 +2409,8 @@ apply_handle_commit_prepared(StringInfo s)
CommitTransactionCommand();
pgstat_report_stat(false);
- store_flush_position(prepare_data.end_lsn, XactLastCommitEnd);
+ store_flush_position(prepare_data.end_lsn, XactLastCommitEnd,
+ InvalidTransactionId);
in_remote_transaction = false;
/*
@@ -2251,7 +2479,8 @@ apply_handle_rollback_prepared(StringInfo s)
* transaction because we always flush the WAL record for it. See
* apply_handle_prepare.
*/
- store_flush_position(rollback_data.rollback_end_lsn, InvalidXLogRecPtr);
+ store_flush_position(rollback_data.rollback_end_lsn, InvalidXLogRecPtr,
+ InvalidTransactionId);
in_remote_transaction = false;
/*
@@ -2313,7 +2542,8 @@ apply_handle_stream_prepare(StringInfo s)
* It is okay not to set the local_end LSN for the prepare because
* we always flush the prepare record. See apply_handle_prepare.
*/
- store_flush_position(prepare_data.end_lsn, InvalidXLogRecPtr);
+ store_flush_position(prepare_data.end_lsn, InvalidXLogRecPtr,
+ InvalidTransactionId);
in_remote_transaction = false;
@@ -2505,8 +2735,17 @@ apply_handle_stream_start(StringInfo s)
set_apply_error_context_xact(stream_xid, InvalidXLogRecPtr);
- /* Try to allocate a worker for the streaming transaction. */
- if (first_segment)
+ /*
+ * Try to allocate a worker for the streaming transaction.
+ *
+ * TODO: Support assigning streaming transactions to parallel apply workers
+ * even when non-streaming transactions are running and better dependency
+ * handling between streaming and non-streaming transactions.
+ */
+ if (first_segment &&
+ am_leader_apply_worker() &&
+ (!TransactionIdIsValid(last_parallelized_remote_xid) ||
+ pa_transaction_committed(last_parallelized_remote_xid)))
pa_allocate_worker(stream_xid, true);
apply_action = get_transaction_apply_action(stream_xid, &winfo);
@@ -3021,8 +3260,7 @@ apply_spooled_messages(FileSet *stream_fileset, TransactionId xid,
int fileno;
pgoff_t offset;
- if (!am_parallel_apply_worker())
- maybe_start_skipping_changes(lsn);
+ maybe_start_skipping_changes(lsn);
/* Make sure we have an open transaction */
begin_replication_step();
@@ -3292,7 +3530,8 @@ apply_handle_commit_internal(LogicalRepCommitData *commit_data)
pgstat_report_stat(false);
- store_flush_position(commit_data->end_lsn, XactLastCommitEnd);
+ store_flush_position(commit_data->end_lsn, XactLastCommitEnd,
+ InvalidTransactionId);
}
else
{
@@ -3317,7 +3556,8 @@ apply_handle_relation(StringInfo s)
{
LogicalRepRelation *rel;
- if (handle_streamed_transaction(LOGICAL_REP_MSG_RELATION, s))
+ if (handle_parallelized_transaction(LOGICAL_REP_MSG_RELATION, s) ||
+ handle_streamed_transaction(LOGICAL_REP_MSG_RELATION, s))
return;
rel = logicalrep_read_rel(s);
@@ -3325,6 +3565,8 @@ apply_handle_relation(StringInfo s)
/* Also reset all entries in the partition map that refer to remoterel. */
logicalrep_partmap_reset_relmap(rel);
+
+ pa_distribute_schema_changes_to_workers(rel);
}
/*
@@ -3340,7 +3582,8 @@ apply_handle_type(StringInfo s)
{
LogicalRepTyp typ;
- if (handle_streamed_transaction(LOGICAL_REP_MSG_TYPE, s))
+ if (handle_parallelized_transaction(LOGICAL_REP_MSG_TYPE, s) ||
+ handle_streamed_transaction(LOGICAL_REP_MSG_TYPE, s))
return;
logicalrep_read_typ(s, &typ);
@@ -3400,6 +3643,7 @@ apply_handle_insert(StringInfo s)
* streamed transactions.
*/
if (is_skipping_changes() ||
+ handle_parallelized_transaction(LOGICAL_REP_MSG_INSERT, s) ||
handle_streamed_transaction(LOGICAL_REP_MSG_INSERT, s))
return;
@@ -3560,6 +3804,7 @@ apply_handle_update(StringInfo s)
* streamed transactions.
*/
if (is_skipping_changes() ||
+ handle_parallelized_transaction(LOGICAL_REP_MSG_UPDATE, s) ||
handle_streamed_transaction(LOGICAL_REP_MSG_UPDATE, s))
return;
@@ -3784,6 +4029,7 @@ apply_handle_delete(StringInfo s)
* streamed transactions.
*/
if (is_skipping_changes() ||
+ handle_parallelized_transaction(LOGICAL_REP_MSG_DELETE, s) ||
handle_streamed_transaction(LOGICAL_REP_MSG_DELETE, s))
return;
@@ -4420,6 +4666,7 @@ apply_handle_truncate(StringInfo s)
* streamed transactions.
*/
if (is_skipping_changes() ||
+ handle_parallelized_transaction(LOGICAL_REP_MSG_TRUNCATE, s) ||
handle_streamed_transaction(LOGICAL_REP_MSG_TRUNCATE, s))
return;
@@ -4650,6 +4897,10 @@ apply_dispatch(StringInfo s)
* check which entries on it are already locally flushed. Those we can report
* as having been flushed.
*
+ * For non-streaming transactions managed by a parallel apply worker, we will
+ * get the local commit end from the shared parallel apply worker info once the
+ * transaction has been committed by the worker.
+ *
* The have_pending_txes is true if there are outstanding transactions that
* need to be flushed.
*/
@@ -4659,6 +4910,7 @@ get_flush_position(XLogRecPtr *write, XLogRecPtr *flush,
{
dlist_mutable_iter iter;
XLogRecPtr local_flush = GetFlushRecPtr(NULL);
+ List *committed_pa_xid = NIL;
*write = InvalidXLogRecPtr;
*flush = InvalidXLogRecPtr;
@@ -4668,6 +4920,40 @@ get_flush_position(XLogRecPtr *write, XLogRecPtr *flush,
FlushPosition *pos =
dlist_container(FlushPosition, node, iter.cur);
+ /*
+ * If the transaction was assigned to a parallel apply worker, attempt
+ * to retrieve its commit end LSN from that worker.
+ */
+ if (TransactionIdIsValid(pos->pa_remote_xid) &&
+ !XLogRecPtrIsValid(pos->local_end))
+ {
+ bool skipped_write;
+
+ pos->local_end = pa_get_last_commit_end(pos->pa_remote_xid, true,
+ &skipped_write);
+
+ elog(DEBUG1,
+ "got commit end from parallel apply worker, "
+ "txn: %u, remote_end %X/%X, local_end %X/%X",
+ pos->pa_remote_xid, LSN_FORMAT_ARGS(pos->remote_end),
+ LSN_FORMAT_ARGS(pos->local_end));
+
+ /*
+ * Break the loop if the worker has not finished applying the
+ * transaction. There's no need to check subsequent transactions,
+ * as they must commit after the current transaction being
+ * examined and thus won't have their commit end available yet.
+ */
+ if (!skipped_write && XLogRecPtrIsInvalid(pos->local_end))
+ break;
+
+ committed_pa_xid = lappend_xid(committed_pa_xid, pos->pa_remote_xid);
+ }
+
+ /*
+ * Worker has finished applying or the transaction was applied in the
+ * leader apply worker.
+ */
*write = pos->remote_end;
if (pos->local_end <= local_flush)
@@ -4676,29 +4962,22 @@ get_flush_position(XLogRecPtr *write, XLogRecPtr *flush,
dlist_delete(iter.cur);
pfree(pos);
}
- else
- {
- /*
- * Don't want to uselessly iterate over the rest of the list which
- * could potentially be long. Instead get the last element and
- * grab the write position from there.
- */
- pos = dlist_tail_element(FlushPosition, node,
- &lsn_mapping);
- *write = pos->remote_end;
- *have_pending_txes = true;
- return;
- }
}
*have_pending_txes = !dlist_is_empty(&lsn_mapping);
+
+ delete_replica_identity_entries_for_txns(committed_pa_xid);
}
/*
* Store current remote/local lsn pair in the tracking list.
+ *
+ * pa_remote_xid should be a valid transaction ID only when the transaction was
+ * assigned to a parallel apply worker; otherwise, pass InvalidTransactionId.
*/
void
-store_flush_position(XLogRecPtr remote_lsn, XLogRecPtr local_lsn)
+store_flush_position(XLogRecPtr remote_lsn, XLogRecPtr local_lsn,
+ TransactionId pa_remote_xid)
{
FlushPosition *flushpos;
@@ -4716,7 +4995,7 @@ store_flush_position(XLogRecPtr remote_lsn, XLogRecPtr local_lsn)
flushpos = palloc_object(FlushPosition);
flushpos->local_end = local_lsn;
flushpos->remote_end = remote_lsn;
- flushpos->pa_remote_xid = InvalidTransactionId;
+ flushpos->pa_remote_xid = pa_remote_xid;
dlist_push_tail(&lsn_mapping, &flushpos->node);
MemoryContextSwitchTo(ApplyMessageContext);
@@ -6841,6 +7120,22 @@ static void
maybe_start_skipping_changes(XLogRecPtr finish_lsn)
{
Assert(!is_skipping_changes());
+
+ /*
+ * For streaming transactions applied in a parallel apply worker, the remote
+ * end LSN is unknown, so skipping is not possible. For non-streaming
+ * transactions, the leader determines whether to skip the transaction. If
+ * skipping is needed, the leader simply does not send the transaction to a
+ * parallel apply worker.
+ */
+ if (am_parallel_apply_worker())
+ return;
+
+ /*
+ * These assertions apply only to leader apply and table sync workers.
+ * Parallel workers may apply spooled BEGIN or STREAM_START messages, which
+ * can set these flags to true.
+ */
Assert(!in_remote_transaction);
Assert(!in_streamed_transaction);
diff --git a/src/include/replication/logicalproto.h b/src/include/replication/logicalproto.h
index cb0a8f440c0..d9f375295de 100644
--- a/src/include/replication/logicalproto.h
+++ b/src/include/replication/logicalproto.h
@@ -259,6 +259,8 @@ extern void logicalrep_write_message(StringInfo out, TransactionId xid, XLogRecP
extern void logicalrep_write_rel(StringInfo out, TransactionId xid,
Relation rel, Bitmapset *columns,
PublishGencolsType include_gencols_type);
+extern void logicalrep_write_internal_rel(StringInfo out,
+ LogicalRepRelation *rel);
extern LogicalRepRelation *logicalrep_read_rel(StringInfo in);
extern void logicalrep_write_typ(StringInfo out, TransactionId xid,
Oid typoid);
diff --git a/src/include/replication/logicalrelation.h b/src/include/replication/logicalrelation.h
index 09890cef8d4..cd852337165 100644
--- a/src/include/replication/logicalrelation.h
+++ b/src/include/replication/logicalrelation.h
@@ -58,6 +58,8 @@ extern void logicalrep_rel_close(LogicalRepRelMapEntry *rel,
LOCKMODE lockmode);
extern bool IsIndexUsableForReplicaIdentityFull(Relation idxrel, AttrMap *attrmap);
extern Oid GetRelationIdentityOrPK(Relation rel);
+extern int logicalrep_get_num_rels(void);
+extern void logicalrep_write_all_rels(StringInfo out);
extern LogicalRepRelMapEntry *logicalrep_get_relentry(LogicalRepRelId remoteid);
#endif /* LOGICALRELATION_H */
diff --git a/src/include/replication/worker_internal.h b/src/include/replication/worker_internal.h
index 3c88d9e1027..02a165ccc0d 100644
--- a/src/include/replication/worker_internal.h
+++ b/src/include/replication/worker_internal.h
@@ -345,7 +345,8 @@ extern void SetupApplyOrSyncWorker(int worker_slot);
extern void DisableSubscriptionAndExit(void);
-extern void store_flush_position(XLogRecPtr remote_lsn, XLogRecPtr local_lsn);
+extern void store_flush_position(XLogRecPtr remote_lsn, XLogRecPtr local_lsn,
+ TransactionId pa_remote_xid);
/* Function for apply error callback */
extern void apply_error_callback(void *arg);
@@ -354,13 +355,15 @@ extern void set_apply_error_context_origin(char *originname);
/* Parallel apply worker setup and interactions */
extern void pa_allocate_worker(TransactionId xid, bool stream_txn);
extern ParallelApplyWorkerInfo *pa_find_worker(TransactionId xid);
-extern XLogRecPtr pa_get_last_commit_end(TransactionId xid, bool *skipped_write);
+extern XLogRecPtr pa_get_last_commit_end(TransactionId xid, bool delete_entry,
+ bool *skipped_write);
extern void pa_detach_all_error_mq(void);
extern void apply_handle_internal_message(StringInfo s);
extern bool pa_send_data(ParallelApplyWorkerInfo *winfo, Size nbytes,
const void *data);
+extern void pa_distribute_schema_changes_to_workers(LogicalRepRelation *rel);
extern void pa_switch_to_partial_serialize(ParallelApplyWorkerInfo *winfo,
bool stream_locked);
@@ -386,8 +389,9 @@ extern void pa_decr_and_wait_stream_block(void);
extern void pa_xact_finish(ParallelApplyWorkerInfo *winfo,
XLogRecPtr remote_lsn);
extern bool pa_transaction_committed(TransactionId xid);
-
+extern void pa_commit_transaction(void);
extern void pa_wait_for_depended_transaction(TransactionId xid);
+extern void pa_add_parallelized_transaction(TransactionId xid);
#define isParallelApplyWorker(worker) ((worker)->in_use && \
(worker)->type == WORKERTYPE_PARALLEL_APPLY)
diff --git a/src/test/subscription/meson.build b/src/test/subscription/meson.build
index fdc04d91f41..48ae698e786 100644
--- a/src/test/subscription/meson.build
+++ b/src/test/subscription/meson.build
@@ -48,7 +48,7 @@ tests += {
't/036_sequences.pl',
't/037_except.pl',
't/038_walsnd_shutdown_timeout.pl',
- 't/039_row_filter_toast_ri.pl',
+ 't/050_parallel_apply.pl',
't/100_bugs.pl',
],
},
diff --git a/src/test/subscription/t/001_rep_changes.pl b/src/test/subscription/t/001_rep_changes.pl
index 7d41715ed81..c863b430bec 100644
--- a/src/test/subscription/t/001_rep_changes.pl
+++ b/src/test/subscription/t/001_rep_changes.pl
@@ -16,6 +16,8 @@ $node_publisher->start;
# Create subscriber node
my $node_subscriber = PostgreSQL::Test::Cluster->new('subscriber');
$node_subscriber->init;
+$node_subscriber->append_conf('postgresql.conf',
+ "max_logical_replication_workers = 10");
$node_subscriber->start;
# Create some preexisting content on publisher
diff --git a/src/test/subscription/t/010_truncate.pl b/src/test/subscription/t/010_truncate.pl
index 945505d0239..e15a6bb2a03 100644
--- a/src/test/subscription/t/010_truncate.pl
+++ b/src/test/subscription/t/010_truncate.pl
@@ -17,7 +17,7 @@ $node_publisher->start;
my $node_subscriber = PostgreSQL::Test::Cluster->new('subscriber');
$node_subscriber->init;
$node_subscriber->append_conf('postgresql.conf',
- qq(max_logical_replication_workers = 6));
+ qq(max_logical_replication_workers = 7));
$node_subscriber->start;
my $publisher_connstr = $node_publisher->connstr . ' dbname=postgres';
diff --git a/src/test/subscription/t/015_stream.pl b/src/test/subscription/t/015_stream.pl
index ac96bc3f009..705bc83f5ea 100644
--- a/src/test/subscription/t/015_stream.pl
+++ b/src/test/subscription/t/015_stream.pl
@@ -232,6 +232,12 @@ $node_subscriber->wait_for_log(
$node_publisher->safe_psql('postgres', "INSERT INTO test_tab_2 values(1)");
+# FIXME: Currently, non-streaming transactions are applied in parallel by
+# default. So, the first transaction is handled by a parallel apply worker. To
+# trigger the deadlock, initiate an more transaction to be applied by the
+# leader.
+$node_publisher->safe_psql('postgres', "INSERT INTO test_tab_2 values(1)");
+
$h->query_safe('COMMIT');
$h->quit;
@@ -247,7 +253,7 @@ $node_publisher->wait_for_catchup($appname);
$result =
$node_subscriber->safe_psql('postgres', "SELECT count(*) FROM test_tab_2");
-is($result, qq(5001), 'data replicated to subscriber after dropping index');
+is($result, qq(5002), 'data replicated to subscriber after dropping index');
# Clean up test data from the environment.
$node_publisher->safe_psql('postgres', "TRUNCATE TABLE test_tab_2");
diff --git a/src/test/subscription/t/026_stats.pl b/src/test/subscription/t/026_stats.pl
index 5d457060a02..911005bde20 100644
--- a/src/test/subscription/t/026_stats.pl
+++ b/src/test/subscription/t/026_stats.pl
@@ -16,6 +16,7 @@ $node_publisher->start;
# Create subscriber node.
my $node_subscriber = PostgreSQL::Test::Cluster->new('subscriber');
$node_subscriber->init;
+$node_subscriber->append_conf('postgresql.conf', "max_logical_replication_workers = 10");
$node_subscriber->start;
diff --git a/src/test/subscription/t/027_nosuperuser.pl b/src/test/subscription/t/027_nosuperuser.pl
index 322f5b4cc6a..fdfc44ac729 100644
--- a/src/test/subscription/t/027_nosuperuser.pl
+++ b/src/test/subscription/t/027_nosuperuser.pl
@@ -86,6 +86,7 @@ $node_publisher = PostgreSQL::Test::Cluster->new('publisher');
$node_subscriber = PostgreSQL::Test::Cluster->new('subscriber');
$node_publisher->init(allows_streaming => 'logical');
$node_subscriber->init;
+$node_subscriber->append_conf('postgresql.conf', "max_logical_replication_workers = 10");
$node_publisher->start;
$node_subscriber->start;
$publisher_connstr = $node_publisher->connstr . ' dbname=postgres';
diff --git a/src/test/subscription/t/050_parallel_apply.pl b/src/test/subscription/t/050_parallel_apply.pl
new file mode 100644
index 00000000000..9d6026470d7
--- /dev/null
+++ b/src/test/subscription/t/050_parallel_apply.pl
@@ -0,0 +1,324 @@
+
+# Copyright (c) 2025, PostgreSQL Global Development Group
+
+# This test verifies that a non-streamed transaction can launch a parallel apply
+# worker, and that dependency tracking and commit order preservation work
+# correctly during parallel apply.
+
+use strict;
+use warnings;
+use PostgreSQL::Test::Cluster;
+use PostgreSQL::Test::Utils;
+use Test::More;
+
+if ($ENV{enable_injection_points} ne 'yes')
+{
+ plan skip_all => 'Injection points not supported by this build';
+}
+
+# Initialize publisher node
+my $node_publisher = PostgreSQL::Test::Cluster->new('publisher');
+$node_publisher->init(allows_streaming => 'logical');
+$node_publisher->start;
+
+# Create tables and insert initial data
+$node_publisher->safe_psql(
+ 'postgres', qq(
+ CREATE TABLE regress_tab (id int PRIMARY KEY, value text);
+
+ CREATE TABLE tab_ri_full (id int, value text);
+ ALTER TABLE tab_ri_full REPLICA IDENTITY FULL;
+ INSERT INTO tab_ri_full VALUES (1, 'test');
+
+ CREATE TABLE tab_toast (a text NOT NULL, b text NOT NULL);
+ ALTER TABLE tab_toast ALTER COLUMN a SET STORAGE EXTERNAL;
+ CREATE UNIQUE INDEX tab_toast_ri_index on tab_toast (a, b);
+ ALTER TABLE tab_toast REPLICA IDENTITY USING INDEX tab_toast_ri_index;
+ INSERT INTO tab_toast(a, b) VALUES(repeat('1234567890', 200), '1234567890');
+));
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO regress_tab VALUES (generate_series(1, 10), 'test');");
+
+# Create a publication
+$node_publisher->safe_psql('postgres',
+ "CREATE PUBLICATION regress_pub FOR ALL TABLES;");
+
+# Initialize subscriber node
+my $node_subscriber = PostgreSQL::Test::Cluster->new('subscriber');
+$node_subscriber->init;
+$node_subscriber->append_conf('postgresql.conf', "log_min_messages = debug1");
+$node_subscriber->append_conf('postgresql.conf',
+ "max_logical_replication_workers = 10");
+$node_subscriber->start;
+
+# Check if the extension injection_points is available, as it may be
+# possible that this script is run with installcheck, where the module
+# would not be installed by default.
+if (!$node_subscriber->check_extension('injection_points'))
+{
+ plan skip_all => 'Extension injection_points not installed';
+}
+
+$node_subscriber->safe_psql('postgres', 'CREATE EXTENSION injection_points;');
+
+# Create a subscription
+my $publisher_connstr = $node_publisher->connstr . ' dbname=postgres';
+
+$node_subscriber->safe_psql(
+ 'postgres', qq(
+ CREATE TABLE regress_tab (id int PRIMARY KEY, value text);
+
+ CREATE TABLE tab_ri_full (id int, value text);
+ ALTER TABLE tab_ri_full REPLICA IDENTITY FULL;
+
+ CREATE TABLE tab_toast (a text NOT NULL, b text NOT NULL);
+ ALTER TABLE tab_toast ALTER COLUMN a SET STORAGE EXTERNAL;
+ CREATE UNIQUE INDEX tab_toast_ri_index on tab_toast (a, b);
+ ALTER TABLE tab_toast REPLICA IDENTITY USING INDEX tab_toast_ri_index;
+));
+$node_subscriber->safe_psql('postgres',
+ "CREATE SUBSCRIPTION regress_sub CONNECTION '$publisher_connstr' PUBLICATION regress_pub;");
+
+# Wait for initial table sync to finish
+$node_subscriber->wait_for_subscription_sync($node_publisher, 'regress_sub');
+
+##################################################
+# Test that a non-streamed transaction can be applied in a parallel apply worker
+##################################################
+
+# Insert tuples on publisher
+#
+# XXX This may not enough to launch a parallel apply worker, because
+# table_states_not_ready is not discarded yet.
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO regress_tab VALUES (generate_series(11, 20), 'test');");
+$node_publisher->wait_for_catchup('regress_sub');
+
+# Insert tuples again
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO regress_tab VALUES (generate_series(21, 30), 'test');");
+$node_publisher->wait_for_catchup('regress_sub');
+
+# Verify the parallel apply worker is launched
+my $result = $node_subscriber->safe_psql('postgres',
+ "SELECT count(1) FROM pg_stat_activity WHERE backend_type = 'logical replication parallel worker'");
+is($result, '1', "parallel apply worker is launched by a non-streamed transaction");
+
+##################################################
+# Test that the basic replica identity dependency tracking and commit order
+# preservation work correctly during parallel apply.
+##################################################
+
+# Attach an injection_point. Parallel workers would wait before the commit
+$node_subscriber->safe_psql('postgres',
+ "SELECT injection_points_attach('parallel-worker-before-commit','wait');"
+);
+
+# Insert tuples on publisher
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO regress_tab VALUES (generate_series(31, 40), 'test');");
+
+# Wait until the parallel worker enters the injection point.
+$node_subscriber->wait_for_event('logical replication parallel worker',
+ 'parallel-worker-before-commit');
+
+my $offset = -s $node_subscriber->logfile;
+
+# Insert tuples on publisher again. This transaction is independent from the
+# previous one, but the parallel worker would wait till it finishes
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO regress_tab VALUES (generate_series(41, 50), 'test');");
+
+# Verify the parallel worker waits for the transaction
+my $str = $node_subscriber->wait_for_log(qr/wait for depended xid ([1-9][0-9]+)/, $offset);
+my $xid = $str =~ /wait for depended xid ([1-9][0-9]+)/;
+
+ok(1, "commit order dependency detected for parallel apply");
+
+$offset = -s $node_subscriber->logfile;
+
+# Update tuples which have not been applied yet on subscriber because the
+# parallel worker stops at the injection point. Newly assigned worker also
+# waits for the same transactions as above.
+$node_publisher->safe_psql('postgres',
+ "UPDATE regress_tab SET value = 'updated' WHERE id BETWEEN 31 AND 35;");
+
+# Verify the dependency is detected for the update
+$node_subscriber->wait_for_log(qr/found conflicting replica identity change on table [1-9][0-9]+ from $xid/, $offset);
+
+# Verify the parallel worker waits for the same transaction
+$node_subscriber->wait_for_log(qr/wait for depended xid $xid/, $offset);
+
+ok(1, "replica identity dependency detected for parallel apply");
+
+# Wakeup the parallel worker. We detach first no to stop other parallel workers
+$node_subscriber->safe_psql('postgres', qq[
+ SELECT injection_points_detach('parallel-worker-before-commit');
+ SELECT injection_points_wakeup('parallel-worker-before-commit');
+]);
+
+# Verify the parallel worker wakes up
+$node_subscriber->wait_for_log(qr/finish waiting for depended xid $xid/, $offset);
+
+$node_publisher->wait_for_catchup('regress_sub');
+
+$result =
+ $node_subscriber->safe_psql('postgres', "SELECT count(1) FROM regress_tab");
+is ($result, 50, 'inserts are replicated to subscriber');
+
+$result =
+ $node_subscriber->safe_psql('postgres',
+ "SELECT count(1) FROM regress_tab WHERE value = 'updated'");
+is ($result, 5, 'updates are also replicated to subscriber');
+
+##################################################
+# Test that the dependency tracking works correctly for unchanged toasted RI
+# columns.
+##################################################
+
+# Attach an injection_point. Parallel workers would wait before the commit
+$node_subscriber->safe_psql('postgres',
+ "SELECT injection_points_attach('parallel-worker-before-commit','wait');"
+);
+
+# Update one replica identity column but keep toasted column unchanged
+$node_publisher->safe_psql('postgres',
+ "UPDATE tab_toast SET b = '1';");
+
+# Wait until the parallel worker enters the injection point.
+$node_subscriber->wait_for_event('logical replication parallel worker',
+ 'parallel-worker-before-commit');
+
+$offset = -s $node_subscriber->logfile;
+
+# Delete the updated row.
+$node_publisher->safe_psql('postgres',
+ "DELETE FROM tab_toast WHERE b = '1';");
+
+# Verify the dependency is detected for the delete
+$str = $node_subscriber->wait_for_log(qr/found conflicting replica identity change on table [1-9][0-9]+ from ([1-9][0-9]+)/, $offset);
+$xid = $str =~ /found conflicting replica identity change on table [1-9][0-9]+ from ([1-9][0-9]+)/;
+
+# Verify the parallel worker waits for the same transaction
+$node_subscriber->wait_for_log(qr/wait for depended xid $xid/, $offset);
+
+ok(1, "replica identity dependency from unchanged toasted column detected for parallel apply");
+
+# Wakeup the parallel worker. We detach first no to stop other parallel workers
+$node_subscriber->safe_psql('postgres', qq[
+ SELECT injection_points_detach('parallel-worker-before-commit');
+ SELECT injection_points_wakeup('parallel-worker-before-commit');
+]);
+
+# Verify the parallel worker wakes up
+$node_subscriber->wait_for_log(qr/finish waiting for depended xid $xid/, $offset);
+
+$node_publisher->wait_for_catchup('regress_sub');
+
+$result =
+ $node_subscriber->safe_psql('postgres', "SELECT count(1) FROM tab_toast");
+is ($result, 0, 'changes are replicated to subscriber');
+
+##################################################
+# Test that dependency tracking still works for REPLICA IDENTITY FULL when
+# new tuple includes NULL key values.
+##################################################
+
+# Attach an injection_point. Parallel workers would wait before the commit
+$node_subscriber->safe_psql('postgres',
+ "SELECT injection_points_attach('parallel-worker-before-commit','wait');"
+);
+
+# Update one row to a non-NULL value and block commit in parallel worker.
+$node_publisher->safe_psql('postgres',
+ "UPDATE tab_ri_full SET value = NULL WHERE id = 1;");
+
+# Wait until the parallel worker enters the injection point.
+$node_subscriber->wait_for_event('logical replication parallel worker',
+ 'parallel-worker-before-commit');
+
+$offset = -s $node_subscriber->logfile;
+
+# This update sets a NULL value under REPLICA IDENTITY FULL. We must still
+# detect dependency on the preceding update of the same row.
+$node_publisher->safe_psql('postgres',
+ "UPDATE tab_ri_full SET value = 'test' WHERE id = 1;");
+
+# Verify the dependency is detected for the update with NULL key value.
+$str = $node_subscriber->wait_for_log(qr/found conflicting replica identity change on table [1-9][0-9]+ from ([1-9][0-9]+)/, $offset);
+$xid = $str =~ /found conflicting replica identity change on table [1-9][0-9]+ from ([1-9][0-9]+)/;
+
+# Verify the parallel worker waits for the same transaction.
+$node_subscriber->wait_for_log(qr/wait for depended xid $xid/, $offset);
+
+ok(1, "replica identity FULL dependency with NULL values detected for parallel apply");
+
+# Wakeup the parallel worker. We detach first no to stop other parallel workers
+$node_subscriber->safe_psql('postgres', qq[
+ SELECT injection_points_detach('parallel-worker-before-commit');
+ SELECT injection_points_wakeup('parallel-worker-before-commit');
+]);
+
+# Verify the parallel worker wakes up.
+$node_subscriber->wait_for_log(qr/finish waiting for depended xid $xid/, $offset);
+
+$node_publisher->wait_for_catchup('regress_sub');
+
+$result =
+ $node_subscriber->safe_psql('postgres',
+ "SELECT count(1) FROM tab_ri_full WHERE id = 1 AND value = 'test'");
+is ($result, 1, 'update is replicated for REPLICA IDENTITY FULL table');
+
+##################################################
+# Test that table level dependency tracking by TRUNCATE work correctly during
+# parallel apply.
+##################################################
+
+# Truncate the data for upcoming tests
+$node_publisher->safe_psql('postgres', "TRUNCATE TABLE regress_tab;");
+$node_publisher->wait_for_catchup('regress_sub');
+
+# Attach an injection_point. Parallel workers would wait before the commit
+$node_subscriber->safe_psql('postgres',
+ "SELECT injection_points_attach('parallel-worker-before-commit','wait');"
+);
+
+# Insert tuples on publisher
+$node_publisher->safe_psql('postgres',
+ "TRUNCATE regress_tab;");
+
+# Wait until the parallel worker enters the injection point.
+$node_subscriber->wait_for_event('logical replication parallel worker',
+ 'parallel-worker-before-commit');
+
+$offset = -s $node_subscriber->logfile;
+
+# Insert a tuple that conflicts with the TRUNCATE operation.
+$node_publisher->safe_psql('postgres',
+ "INSERT INTO regress_tab VALUES (1, 'test');");
+
+# Verify the dependency is detected for the insert
+$str = $node_subscriber->wait_for_log(qr/found table-wide change affecting [1-9][0-9]+ from ([1-9][0-9]+)/, $offset);
+$xid = $str =~ /found table-wide change affecting [1-9][0-9]+ from ([1-9][0-9]+)/;
+
+ok(1, "table-wide dependency from TRUNCATE detected for parallel apply");
+
+# Wakeup the parallel worker. We detach first no to stop other parallel workers
+$node_subscriber->safe_psql('postgres', qq[
+ SELECT injection_points_detach('parallel-worker-before-commit');
+ SELECT injection_points_wakeup('parallel-worker-before-commit');
+]);
+
+# Verify the parallel worker waits for the same transaction
+$node_subscriber->wait_for_log(qr/wait for depended xid $xid/, $offset);
+
+# Verify the parallel worker wakes up
+$node_subscriber->wait_for_log(qr/finish waiting for depended xid $xid/, $offset);
+
+$node_publisher->wait_for_catchup('regress_sub');
+
+$result =
+ $node_subscriber->safe_psql('postgres', "SELECT count(1) FROM regress_tab");
+is ($result, 1, 'inserts are replicated to subscriber');
+
+done_testing();
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 9609fd05967..c9b24dbfa59 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -2164,6 +2164,7 @@ ParallelHashGrowth
ParallelHashJoinBatch
ParallelHashJoinBatchAccessor
ParallelHashJoinState
+ParallelizedTxnEntry
ParallelIndexScanDesc
ParallelizedTxnEntry
ParallelSlot
@@ -4263,6 +4264,7 @@ rendezvousHashEntry
rep
replace_rte_variables_callback
replace_rte_variables_context
+replica_identity_hash
report_error_fn
ret_type
rewind_source
--
2.43.0
[application/octet-stream] v21-0002-Introduce-internal-messages-to-track-dependencie.patch (12.7K, ../TY4PR01MB177180394D8E9C6D263E117E494FE2@TY4PR01MB17718.jpnprd01.prod.outlook.com/10-v21-0002-Introduce-internal-messages-to-track-dependencie.patch)
download | inline diff:
From 45a6e95e81cc275be655e4006112b650bc4cf63d Mon Sep 17 00:00:00 2001
From: Zhijie Hou <[email protected]>
Date: Wed, 29 Apr 2026 16:41:44 +0800
Subject: [PATCH v21 2/9] Introduce internal messages to track dependencies
This patch introduces a new set of internal worker message types (see
PAWorkerMsgType). These types of messages are generated by a leader worker
and sent to parallel apply workers based on the needs. For now, two types of
messages exist: PA_MSG_XACT_DEPENDENCY and PA_MSG_RELMAP.
PA_MSG_XACT_DEPENDENCY ensures that dependent transactions are
committed in the correct order. It has a list of transaction IDs that parallel
workers must wait for. This type of message is generated when the leader
detects a dependency between the current and other transactions, or just before
the COMMIT message. The latter one is used to preserve the commit ordering
between the publisher and the subscriber.
PA_MSG_RELMAP is used to synchronize the remote relation
information between the leader and parallel workers. It has a list of relations
that the leader already knows, and parallel workers also update the relmap in
response to the message. This type of message is generated when the leader
allocates a new parallel worker to the transaction, or when the publisher sends
additional RELATION messages. This synchronization is necessary for parallel
apply workers to map local replication target relations to their remote
counterparts during change application. Since the walsender does not send remote
relation information with every transaction, the parallel apply worker may not
have up-to-date relation info unless synchronized by the leader.
In the logical replication protocol, the above messages are encapsulated within
the LOGICAL_REP_MSG_INTERNAL_MESSAGE type. Later patches will use a uniform
format (LOGICAL_REP_MSG_INTERNAL_MESSAGE + PAWorkerMsgType + internal data) for
sending to parallel apply workers or serializing to disk.
Author: Zhijie Hou <[email protected]>
Author: Hayato Kuroda <[email protected]>
---
.../replication/logical/applyparallelworker.c | 187 ++++++++++++++++--
src/backend/replication/logical/proto.c | 2 +
src/backend/replication/logical/worker.c | 4 +
src/include/replication/logicalproto.h | 8 +
src/include/replication/worker_internal.h | 17 ++
src/tools/pgindent/typedefs.list | 1 +
6 files changed, 202 insertions(+), 17 deletions(-)
diff --git a/src/backend/replication/logical/applyparallelworker.c b/src/backend/replication/logical/applyparallelworker.c
index 5eb0679e850..efde9181a34 100644
--- a/src/backend/replication/logical/applyparallelworker.c
+++ b/src/backend/replication/logical/applyparallelworker.c
@@ -772,6 +772,73 @@ ProcessParallelApplyInterrupts(void)
}
}
+/*
+ * Handle internal dependency information.
+ *
+ * Wait for all transactions listed in the message to commit.
+ */
+static void
+apply_handle_internal_dependency(StringInfo s)
+{
+ int nxids = pq_getmsgint(s, 4);
+
+ for (int i = 0; i < nxids; i++)
+ {
+ TransactionId xid = pq_getmsgint(s, 4);
+
+ pa_wait_for_depended_transaction(xid);
+ }
+}
+
+/*
+ * Handle internal relation information.
+ *
+ * Update all relation details in the relation map cache.
+ */
+static void
+apply_handle_internal_relation(StringInfo s)
+{
+ int nrels = pq_getmsgint(s, 4);
+
+ for (int i = 0; i < nrels; i++)
+ {
+ LogicalRepRelation *rel = logicalrep_read_rel(s);
+
+ logicalrep_relmap_update(rel);
+
+ /* Also reset all entries in the partition map that refer to remoterel. */
+ logicalrep_partmap_reset_relmap(rel);
+
+ elog(DEBUG1, "parallel apply worker init relmap for %s",
+ rel->relname);
+ }
+}
+
+/*
+ * Handle an internal message generated by the leader apply worker.
+ */
+void
+apply_handle_internal_message(StringInfo s)
+{
+ PAWorkerMsgType action = pq_getmsgbyte(s);
+
+ Assert(am_parallel_apply_worker());
+
+ switch (action)
+ {
+ case PA_MSG_XACT_DEPENDENCY:
+ apply_handle_internal_dependency(s);
+ break;
+ case PA_MSG_RELMAP:
+ apply_handle_internal_relation(s);
+ break;
+ default:
+ ereport(ERROR,
+ (errcode(ERRCODE_PROTOCOL_VIOLATION),
+ errmsg("invalid worker internal message type \"??? (%d)\"", action)));
+ }
+}
+
/* Parallel apply worker main loop. */
static void
LogicalParallelApplyLoop(shm_mq_handle *mqh)
@@ -780,6 +847,14 @@ LogicalParallelApplyLoop(shm_mq_handle *mqh)
ErrorContextCallback errcallback;
MemoryContext oldcxt = CurrentMemoryContext;
+ /*
+ * Ensure LOGICAL_REP_MSG_INTERNAL_MESSAGE does not conflict with
+ * PqReplMsg_WALData ('d'), as parallel apply workers may receive both types
+ * of messages.
+ */
+ StaticAssertDecl(LOGICAL_REP_MSG_INTERNAL_MESSAGE != PqReplMsg_WALData,
+ "LOGICAL_REP_MSG_INTERNAL_MESSAGE conflicts with PqReplMsg_WALData");
+
/*
* Init the ApplyMessageContext which we clean up after each replication
* protocol message.
@@ -818,26 +893,46 @@ LogicalParallelApplyLoop(shm_mq_handle *mqh)
initReadOnlyStringInfo(&s, data, len);
- /*
- * The first byte of messages sent from leader apply worker to
- * parallel apply workers can only be PqReplMsg_WALData.
- */
c = pq_getmsgbyte(&s);
- if (c != PqReplMsg_WALData)
- elog(ERROR, "unexpected message \"%c\"", c);
- /*
- * Ignore statistics fields that have been updated by the leader
- * apply worker.
- *
- * XXX We can avoid sending the statistics fields from the leader
- * apply worker but for that, it needs to rebuild the entire
- * message by removing these fields which could be more work than
- * simply ignoring these fields in the parallel apply worker.
- */
- s.cursor += SIZE_STATS_MESSAGE;
+ if (c == PqReplMsg_WALData)
+ {
+ /*
+ * Ignore statistics fields that have been updated by the
+ * leader apply worker.
+ *
+ * XXX We can avoid sending the statistics fields from the
+ * leader apply worker but for that, it needs to rebuild the
+ * entire message by removing these fields which could be more
+ * work than simply ignoring these fields in the parallel apply
+ * worker.
+ */
+ s.cursor += SIZE_STATS_MESSAGE;
+
+ apply_dispatch(&s);
+ }
+ else if (c == LOGICAL_REP_MSG_INTERNAL_MESSAGE)
+ {
+ /*
+ * Rewind the cursor so that apply_dispatch can re-read the
+ * first byte (LOGICAL_REP_MSG_INTERNAL_MESSAGE) to correctly
+ * handle the internal message. Alternatively, we could call
+ * apply_handle_internal_message directly, but using
+ * apply_dispatch uniformly across all message types is cleaner
+ * and more consistent.
+ */
+ s.cursor--;
- apply_dispatch(&s);
+ apply_dispatch(&s);
+ }
+ else
+ {
+ /*
+ * The first byte of messages sent from leader apply worker to
+ * parallel apply workers can only be 'w' or 'i'.
+ */
+ elog(ERROR, "unexpected message \"%c\"", c);
+ }
}
else if (shmq_res == SHM_MQ_WOULD_BLOCK)
{
@@ -1744,3 +1839,61 @@ pa_attach_parallelized_txn_hash(dsa_handle *pa_dsa_handle,
MemoryContextSwitchTo(oldctx);
}
+
+/*
+ * Wait for the given remote transaction to finish applying by a parallel apply
+ * worker.
+ *
+ * Both leader and parallel apply workers can call this function to wait for a
+ * parallelized transaction to finish.
+ */
+void
+pa_wait_for_depended_transaction(TransactionId xid)
+{
+ /*
+ * Quick exit if parallelized_txns has not been initialized yet. This can
+ * happen when the leader worker calls this function before any parallel
+ * apply workers have been launched.
+ */
+ if (!parallelized_txns)
+ return;
+
+ elog(DEBUG1, "wait for depended xid %u", xid);
+
+ for (;;)
+ {
+ ParallelizedTxnEntry *txn_entry;
+
+ txn_entry = dshash_find(parallelized_txns, &xid, false);
+
+ /* The entry is removed only if the transaction is committed */
+ if (txn_entry == NULL)
+ break;
+
+ dshash_release_lock(parallelized_txns, txn_entry);
+
+ /*
+ * Wait for the parallel apply worker processing the given remote
+ * transaction to finish applying and release its lock.
+ */
+ pa_lock_transaction(xid, AccessShareLock);
+ pa_unlock_transaction(xid, AccessShareLock);
+
+ CHECK_FOR_INTERRUPTS();
+
+ /*
+ * Acquiring the lock successfully does not guarantee we can proceed.
+ * The worker may have errored out and released the lock while leaving
+ * its shared hash entry intact, or it may not have acquired the lock
+ * yet because it hasn't processed the BEGIN message. In either case, we
+ * must continue waiting in the loop until the parallel apply worker
+ * finishes applying the transaction, or until the leader notifies us of
+ * a failure and restarts all workers.
+ *
+ * The above race window is small and infrequent, so no WaitLatch is
+ * added.
+ */
+ }
+
+ elog(DEBUG1, "finish waiting for depended xid %u", xid);
+}
diff --git a/src/backend/replication/logical/proto.c b/src/backend/replication/logical/proto.c
index 86ad97cd937..a0cac05cdcd 100644
--- a/src/backend/replication/logical/proto.c
+++ b/src/backend/replication/logical/proto.c
@@ -1253,6 +1253,8 @@ logicalrep_message_type(LogicalRepMsgType action)
return "STREAM ABORT";
case LOGICAL_REP_MSG_STREAM_PREPARE:
return "STREAM PREPARE";
+ case LOGICAL_REP_MSG_INTERNAL_MESSAGE:
+ return "INTERNAL MESSAGE";
}
/*
diff --git a/src/backend/replication/logical/worker.c b/src/backend/replication/logical/worker.c
index 7799266c614..bd53034cf69 100644
--- a/src/backend/replication/logical/worker.c
+++ b/src/backend/replication/logical/worker.c
@@ -3890,6 +3890,10 @@ apply_dispatch(StringInfo s)
apply_handle_stream_prepare(s);
break;
+ case LOGICAL_REP_MSG_INTERNAL_MESSAGE:
+ apply_handle_internal_message(s);
+ break;
+
default:
ereport(ERROR,
(errcode(ERRCODE_PROTOCOL_VIOLATION),
diff --git a/src/include/replication/logicalproto.h b/src/include/replication/logicalproto.h
index 058a955e20c..cb0a8f440c0 100644
--- a/src/include/replication/logicalproto.h
+++ b/src/include/replication/logicalproto.h
@@ -53,6 +53,13 @@
* in logical replication protocol, which uses a single byte to identify a
* message type. Hence the values should be single-byte wide and preferably
* human-readable characters.
+ *
+ * LOGICAL_REP_MSG_INTERNAL_MESSAGE ('i') is reserved for internal messages (see
+ * PAWorkerMsgType for sub-message types) sent from the leader apply
+ * worker to parallel apply workers. Centralizing this definition here allows
+ * all message types to be handled together and avoids the maintenance burden of
+ * ensuring sub-message types do not conflict with regular LogicalRepMsgType
+ * values.
*/
typedef enum LogicalRepMsgType
{
@@ -75,6 +82,7 @@ typedef enum LogicalRepMsgType
LOGICAL_REP_MSG_STREAM_COMMIT = 'c',
LOGICAL_REP_MSG_STREAM_ABORT = 'A',
LOGICAL_REP_MSG_STREAM_PREPARE = 'p',
+ LOGICAL_REP_MSG_INTERNAL_MESSAGE = 'i',
} LogicalRepMsgType;
/*
diff --git a/src/include/replication/worker_internal.h b/src/include/replication/worker_internal.h
index e60e0c3d6d7..43dcae43a62 100644
--- a/src/include/replication/worker_internal.h
+++ b/src/include/replication/worker_internal.h
@@ -242,6 +242,19 @@ typedef struct ParallelApplyWorkerInfo
ParallelApplyWorkerShared *shared;
} ParallelApplyWorkerInfo;
+/*
+ * Parallel apply worker internal message types.
+ *
+ * These types of messages are generated by the leader apply worker and sent to
+ * the parallel apply worker, encapsulated within the
+ * LOGICAL_REP_MSG_INTERNAL_MESSAGE type.
+ */
+typedef enum PAWorkerMsgType
+{
+ PA_MSG_XACT_DEPENDENCY = 'd',
+ PA_MSG_RELMAP = 'r',
+} PAWorkerMsgType;
+
/* Main memory context for apply worker. Permanent during worker lifetime. */
extern PGDLLIMPORT MemoryContext ApplyContext;
@@ -341,6 +354,8 @@ extern void pa_allocate_worker(TransactionId xid);
extern ParallelApplyWorkerInfo *pa_find_worker(TransactionId xid);
extern void pa_detach_all_error_mq(void);
+extern void apply_handle_internal_message(StringInfo s);
+
extern bool pa_send_data(ParallelApplyWorkerInfo *winfo, Size nbytes,
const void *data);
extern void pa_switch_to_partial_serialize(ParallelApplyWorkerInfo *winfo,
@@ -368,6 +383,8 @@ extern void pa_decr_and_wait_stream_block(void);
extern void pa_xact_finish(ParallelApplyWorkerInfo *winfo,
XLogRecPtr remote_lsn);
+extern void pa_wait_for_depended_transaction(TransactionId xid);
+
#define isParallelApplyWorker(worker) ((worker)->in_use && \
(worker)->type == WORKERTYPE_PARALLEL_APPLY)
#define isTableSyncWorker(worker) ((worker)->in_use && \
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 32524990758..0d5959aadb7 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -1929,6 +1929,7 @@ OverridingKind
PACE_HEADER
PACL
PATH
+PAWorkerMsgType
PCtxtHandle
PERL_CONTEXT
PERL_SI
--
2.43.0
view thread (7+ messages)
reply
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Reply to all the recipients using the --to and --cc options:
reply via email
To: [email protected]
Cc: [email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected]
Subject: RE: Parallel Apply
In-Reply-To: <TY4PR01MB177180394D8E9C6D263E117E494FE2@TY4PR01MB17718.jpnprd01.prod.outlook.com>
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
This inbox is served by agora; see mirroring instructions
for how to clone and mirror all data and code used for this inbox