Movatterモバイル変換

[0]ホーム
URL:

Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

 Sign up
Appearance settings

Commit3fabed0

Browse files
committed
Correctly handle relcache invalidation corner case during logical decoding.
When using a historic snapshot for logical decoding it can validlyhappen that a relation that's in the relcache isn't visible to thathistoric snapshot. E.g. if a newly created relation is referenced inthe query that uses the SQL interface for logical decoding and asinval reset occurs.The earlier commit that fixed the error handling for that corner casealready improves the situation as a ERROR is better than hitting anassertion... But it's obviously not good enough. So additionallyallow that case without an error if a historic snapshot is set up -that won't allow an invalid entry to stay in the cache because it's a)already marked invalid and will thus be rebuilt during the next accessb) the syscaches will be reset at the end of decoding.There might be prettier solutions to handle this case, but all that wecould think of so far end up being much more complex than this quitesimple fix.This fixes the assertion failures reported by the buildfarm (markhor,tick, leech) after the introduction of new regression tests in89fd41b. The failure there weren't actually directly caused byCLOBBER_CACHE_ALWAYS but the extraordinary long runtimes due to itlead to sinval resets triggering the behaviour.Discussion: 22459.1418656530@sss.pgh.pa.usBackpatch to 9.4 where logical decoding was introduced.
1 parent31912d0 commit3fabed0

File tree

1 file changed

+12
-0
lines changed

1 file changed

+12
-0
lines changed

‎src/backend/utils/cache/relcache.c

Lines changed: 12 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -2192,6 +2192,18 @@ RelationClearRelation(Relation relation, bool rebuild)
21922192
newrel=RelationBuildDesc(save_relid, false);
21932193
if (newrel==NULL)
21942194
{
2195+
/*
2196+
* We can validly get here, if we're using a historic snapshot in
2197+
* which a relation, accessed from outside logical decoding, is
2198+
* still invisible. In that case it's fine to just mark the
2199+
* relation as invalid and return - it'll fully get reloaded by
2200+
* the cache reset at the end of logical decoding (or at the next
2201+
* access). During normal processing we don't want to ignore this
2202+
* case as it shouldn't happen there, as explained below.
2203+
*/
2204+
if (HistoricSnapshotActive())
2205+
return;
2206+
21952207
/*
21962208
* This shouldn't happen as dropping a relation is intended to be
21972209
* impossible if still referenced (c.f. CheckTableNotInUse()). But

0 commit comments

Comments
 (0)
[8]ページ先頭
©2009-2025 Movatter.jp