Make pg_replication_origin_drop safe against concurrent drops.
authorAmit Kapila <akapila@postgresql.org>
Wed, 10 Feb 2021 01:47:09 +0000 (07:17 +0530)
committerAmit Kapila <akapila@postgresql.org>
Wed, 10 Feb 2021 01:47:09 +0000 (07:17 +0530)
commitcd142e032ebd50ec7974b3633269477c2c72f1cc
treee1ad3b108bf91b5e0ad82f2455ebfec98ad83e93
parent31c7fb41e26bf03dae231c7165a1a16388b2e366
Make pg_replication_origin_drop safe against concurrent drops.

Currently, we get the origin id from the name and then drop the origin by
taking ExclusiveLock on ReplicationOriginRelationId. So, two concurrent
sessions can get the id from the name at the same time and then when they
try to drop the origin, one of the sessions will get the either
"tuple concurrently deleted" or "cache lookup failed for replication
origin ..".

To prevent this race condition we do the entire operation under lock. This
obviates the need for replorigin_drop() API and we have removed it so if
any extension authors are using it they need to instead use
replorigin_drop_by_name. See it's usage in pg_replication_origin_drop().

Author: Peter Smith
Reviewed-by: Amit Kapila, Euler Taveira, Petr Jelinek, and Alvaro
Herrera
Discussion: https://www.postgresql.org/message-id/CAHut%2BPuW8DWV5fskkMWWMqzt-x7RPcNQOtJQBp6SdwyRghCk7A%40mail.gmail.com
src/backend/commands/subscriptioncmds.c
src/backend/replication/logical/origin.c
src/include/replication/origin.h