database/postgres
1
0
Fork 0
mirror of https://github.com/postgres/postgres.git synced 2025-07-31 22:04:40 +03:00
Code Activity

Make the tablesync worker's replication origin drop logic robust.

Browse Source 
In commit f6c5edb8ab, we started to drop the replication origin slots
before tablesync worker exits to avoid consuming more slots than required.
We were dropping the replication origin in the same transaction where we
were marking the tablesync state as SYNCDONE. Now, if there is any error
after we have dropped the origin but before we commit the containing
transaction, the in-memory state of replication progress won't be rolled
back. Due to this, after the restart, tablesync worker can start streaming
from the wrong location and can apply the already processed transaction.

To fix this, we need to opportunistically drop the origin after marking
the tablesync state as SYNCDONE. Even, if the tablesync worker fails to
remove the replication origin before exit, the apply worker ensures to
clean it up afterward.

Reported by Tom Lane as per buildfarm.
Diagnosed-by: Masahiko Sawada
Author: Hou Zhijie
Reviewed-By: Masahiko Sawada, Amit Kapila
Discussion: https://postgr.es/m/20220714115155.GA5439@depesz.com
Discussion: https://postgr.es/m/CAD21AoAw0Oofi4kiDpJBOwpYyBBBkJj=sLUOn4Gd2GjUAKG-fw@mail.gmail.com
This commit is contained in:
Amit Kapila
2022-09-12 12:40:57 +05:30
parent 5015e1e1b5
commit 88f488319b
2 changed files with 75 additions and 41 deletions
Download Patch File Download Diff File Expand all files Collapse all files

25
src/backend/commands/subscriptioncmds.c
View File

@ -931,10 +931,10 @@ AlterSubscription_refresh(Subscription *sub, bool copy_data,
logicalrep_worker_stop(sub->oid, relid);
/*
* For READY state and SYNCDONE state, we would have already
* dropped the tablesync origin.
* For READY state, we would have already dropped the
* tablesync origin.
*/
if (state != SUBREL_STATE_READY && state != SUBREL_STATE_SYNCDONE)
if (state != SUBREL_STATE_READY)
{
char originname[NAMEDATALEN];
@ -942,8 +942,11 @@ AlterSubscription_refresh(Subscription *sub, bool copy_data,
* Drop the tablesync's origin tracking if exists.
*
* It is possible that the origin is not yet created for
* tablesync worker so passing missing_ok = true. This can
* happen for the states before SUBREL_STATE_FINISHEDCOPY.
* tablesync worker, this can happen for the states before
* SUBREL_STATE_FINISHEDCOPY. The tablesync worker or
* apply worker can also concurrently try to drop the
* origin and by this time the origin might be already
* removed. For these reasons, passing missing_ok = true.
*/
ReplicationOriginNameForTablesync(sub->oid, relid, originname,
sizeof(originname));
@ -1516,19 +1519,13 @@ DropSubscription(DropSubscriptionStmt *stmt, bool isTopLevel)
/*
* Drop the tablesync's origin tracking if exists.
*
* For SYNCDONE/READY states, the tablesync origin tracking is known
* to have already been dropped by the tablesync worker.
*
* It is possible that the origin is not yet created for tablesync
* worker so passing missing_ok = true. This can happen for the states
* before SUBREL_STATE_FINISHEDCOPY.
*/
if (rstate->state != SUBREL_STATE_SYNCDONE)
{
ReplicationOriginNameForTablesync(subid, relid, originname,
sizeof(originname));
replorigin_drop_by_name(originname, true, false);
}
ReplicationOriginNameForTablesync(subid, relid, originname,
sizeof(originname));
replorigin_drop_by_name(originname, true, false);
}
/* Clean up dependencies */