Commit9bc1eee

Amit Kapila

committed

Another try to fix BF failure introduced in commitddd5f4f.

Before attempting to sync the slot on standby bypg_sync_replication_slots(), ensure that on the primary restart_lsn forthe slot has moved to a recent WAL position, by re-creating thesubscription and the logical slot.Author: Hou ZhijieDiscussion:https://postgr.es/m/CAA4eK1+d5Lne8vCAn0un4SP9x-ZBr2-xfxg01uSfeBTSCKFZoQ@mail.gmail.com

1 parent0c444a7 commit9bc1eeeCopy full SHA for 9bc1eee

File tree

-10

lines changed

-10

lines changed

Lines changed: 7 additions & 10 deletions

Original file line number	Diff line number	Diff line change
`@@ -223,17 +223,14 @@`
`223`	`223`	`$standby1->append_conf('postgresql.conf','max_slot_wal_keep_size = -1');`
`224`	`224`	`$standby1->reload;`
`225`	`225`
`226`		`-# Enable the subscription to let it catch up to the latest wal position`
`227`		`-$subscriber1->safe_psql('postgres',`
`228`		`-"ALTER SUBSCRIPTION regress_mysub1 ENABLE");`
	`226`	`+# To ensure that restart_lsn has moved to a recent WAL position, we re-create`
	`227`	`+# the subscription and the logical slot.`
	`228`	`+$subscriber1->safe_psql(`
	`229`	`+'postgres',qq[`
	`230`	`+DROP SUBSCRIPTION regress_mysub1;`
	`231`	`+CREATE SUBSCRIPTION regress_mysub1 CONNECTION '$publisher_connstr' PUBLICATION regress_mypub WITH (slot_name = lsub1_slot, copy_data = false, failover = true);`
	`232`	`+]);`
`229`	`233`
`230`		`-# This wait ensures that confirmed_flush_lsn has been moved to latest`
`231`		`-# position.`
`232`		`-$primary->wait_for_catchup('regress_mysub1');`
`233`		`-`
`234`		`-# To ensure that restart_lsn has moved to a recent WAL position, we need`
`235`		`-# to log XLOG_RUNNING_XACTS and make sure the same is processed as well`
`236`		`-$primary->psql('postgres',"CHECKPOINT");`
`237`	`234`	`$primary->wait_for_catchup('regress_mysub1');`
`238`	`235`
`239`	`236`	`# Do not allow any further advancement of the restart_lsn for the lsub1_slot.`

Comments

(0)