From 3641c20b2f5fec59642bb8515cf017f8188c8058 Mon Sep 17 00:00:00 2001 From: Khanna Date: Tue, 24 Dec 2024 12:27:27 +0530 Subject: [PATCH v4] Validate-max_slot_wal_keep_size-in-pg_createsubscriber This patch introduces validation for the 'max_slot_wal_keep_size' GUC in the 'pg_createsubscriber' utility. The utility now checks that the publisher's 'max_slot_wal_keep_size' is set to '-1' during subscription creation. This ensures proper functioning of logical replication slots. The 'pg_createsubscriber' utility is updated to fetch and validate the 'max_slot_wal_keep_size' setting from the publisher. A warning is raised during the '--dry-run' mode if the configuration is set to a non-default value. By ensuring 'max_slot_wal_keep_size' is -1, this patch prevents the potential deletion of required WAL files on the publisher that could disrupt logical replication. A test case has been added to validate correct warning reporting when 'max_slot_wal_keep_size' is misconfigured. --- doc/src/sgml/ref/pg_createsubscriber.sgml | 7 ++++++ src/bin/pg_basebackup/pg_createsubscriber.c | 25 ++++++++++++++++++- .../t/040_pg_createsubscriber.pl | 23 ++++++++++++++--- 3 files changed, 51 insertions(+), 4 deletions(-) diff --git a/doc/src/sgml/ref/pg_createsubscriber.sgml b/doc/src/sgml/ref/pg_createsubscriber.sgml index 26b8e64a4e..59bc97728e 100644 --- a/doc/src/sgml/ref/pg_createsubscriber.sgml +++ b/doc/src/sgml/ref/pg_createsubscriber.sgml @@ -313,6 +313,13 @@ PostgreSQL documentation linkend="guc-max-wal-senders"/> configured to a value greater than or equal to the number of specified databases and existing WAL sender processes. + + + The 'max_slot_wal_keep_size' must be set to -1 to prevent the automatic + removal of WAL files needed by replication slots. Setting this parameter to + a specific size may lead to replication failures if required WAL files are + prematurely deleted. + diff --git a/src/bin/pg_basebackup/pg_createsubscriber.c b/src/bin/pg_basebackup/pg_createsubscriber.c index e96370a9ec..3703dae33a 100644 --- a/src/bin/pg_basebackup/pg_createsubscriber.c +++ b/src/bin/pg_basebackup/pg_createsubscriber.c @@ -849,6 +849,7 @@ check_publisher(const struct LogicalRepInfo *dbinfo) int max_walsenders; int cur_walsenders; int max_prepared_transactions; + int max_slot_wal_keep_size; pg_log_info("checking settings on publisher"); @@ -872,6 +873,7 @@ check_publisher(const struct LogicalRepInfo *dbinfo) * - wal_level = logical * - max_replication_slots >= current + number of dbs to be converted * - max_wal_senders >= current + number of dbs to be converted + * - max_slot_wal_keep_size = -1 (to prevent deletion of required WAL files) * ----------------------------------------------------------------------- */ res = PQexec(conn, @@ -880,7 +882,8 @@ check_publisher(const struct LogicalRepInfo *dbinfo) " (SELECT count(*) FROM pg_catalog.pg_replication_slots)," " pg_catalog.current_setting('max_wal_senders')," " (SELECT count(*) FROM pg_catalog.pg_stat_activity WHERE backend_type = 'walsender')," - " pg_catalog.current_setting('max_prepared_transactions')"); + " pg_catalog.current_setting('max_prepared_transactions')," + " pg_catalog.current_setting('max_slot_wal_keep_size')"); if (PQresultStatus(res) != PGRES_TUPLES_OK) { @@ -895,6 +898,7 @@ check_publisher(const struct LogicalRepInfo *dbinfo) max_walsenders = atoi(PQgetvalue(res, 0, 3)); cur_walsenders = atoi(PQgetvalue(res, 0, 4)); max_prepared_transactions = atoi(PQgetvalue(res, 0, 5)); + max_slot_wal_keep_size = atoi(PQgetvalue(res, 0, 6)); PQclear(res); @@ -905,6 +909,8 @@ check_publisher(const struct LogicalRepInfo *dbinfo) pg_log_debug("publisher: current wal senders: %d", cur_walsenders); pg_log_debug("publisher: max_prepared_transactions: %d", max_prepared_transactions); + pg_log_debug("publisher: max_slot_wal_keep_size: %d", + max_slot_wal_keep_size); disconnect_database(conn, false); @@ -939,6 +945,23 @@ check_publisher(const struct LogicalRepInfo *dbinfo) "Prepared transactions will be replicated at COMMIT PREPARED."); } +/* + * Validate max_slot_wal_keep_size + * Logical replication requires max_slot_wal_keep_size to be set to -1 on the + * publisher to prevent the deletion of WAL files that are still needed by + * replication slots. If this parameter is set to a non-default value, it may + * cause replication failures due to required WAL files being prematurely + * removed. + */ + if (dry_run && max_slot_wal_keep_size != -1) + { + pg_log_error("publisher requires max_slot_wal_keep_size to be -1, but only %d remain", + max_slot_wal_keep_size); + pg_log_error_hint("Change the configuration parameter \"%s\" on the publisher to %d.", + "max_slot_wal_keep_size", -1); + failed = true; + } + pg_free(wal_level); if (failed) diff --git a/src/bin/pg_basebackup/t/040_pg_createsubscriber.pl b/src/bin/pg_basebackup/t/040_pg_createsubscriber.pl index 0a900edb65..f4c1e079b1 100644 --- a/src/bin/pg_basebackup/t/040_pg_createsubscriber.pl +++ b/src/bin/pg_basebackup/t/040_pg_createsubscriber.pl @@ -318,8 +318,14 @@ $node_p->safe_psql($db1, $node_p->wait_for_replay_catchup($node_s); $node_s->stop; -# dry run mode on node S -command_ok( +# Configure 'max_slot_wal_keep_size = 1' on the publisher and +# reload configuration +$node_p->append_conf('postgresql.conf', 'max_slot_wal_keep_size = 1'); +$node_p->reload; + +# dry run mode on node S and verify the error message for misconfigured +# 'max_slot_wal_keep_size' +command_checks_all( [ 'pg_createsubscriber', '--verbose', '--recovery-timeout', "$PostgreSQL::Test::Utils::timeout_default", @@ -335,7 +341,18 @@ command_ok( $db1, '--database', $db2 ], - 'run pg_createsubscriber --dry-run on node S'); + 1, + [qr/./], + [ + qr/pg_createsubscriber: error: publisher requires max_slot_wal_keep_size to be -1/, + qr/Change the configuration parameter "max_slot_wal_keep_size" on the publisher to -1./, + ], + 'Validate error for misconfigured max_slot_wal_keep_size on the publisher' +); + +# Reset 'max_slot_wal_keep_size' to default after the test +$node_p->append_conf('postgresql.conf', 'max_slot_wal_keep_size = -1'); +$node_p->reload; # Check if node S is still a standby $node_s->start; -- 2.41.0.windows.3