From 0230da90f7d930422db388353a4040ee4afd6112 Mon Sep 17 00:00:00 2001 From: wuchengwen Date: Sun, 8 Sep 2024 12:09:12 +0800 Subject: [PATCH] Fix orphaned 2pc file which may casue instance restart failed There is a race condition for global transaction when removing 2pc files: 1. gxact1 is doing `FinishPreparedTransaction` and checkpoint is issued, so gxact1 will generate a 2pc file. 2. then gxact1 was removed from `TwoPhaseState->prepXacts` and was put into freelist. 3. but just before gxact1 remove its 2pc file, gxact2 is issued, gxact2 will reuse the same state memory of gxact1 and will reset `gxact->ondisk` to false. 4. gxact1 continue and found that `gxact->ondisk` is false, it won't remove its 2pc file. This file is orphaned. If gxact1's local xid is not frozen, the startup process will remove the orphaned 2pc file. However, if the clog file is truncated by `vacuum`, the startup process will raise error 'could not access status of transaction xxx', due to it could not found the transaction's status file in dir `pg_xact`. --- src/backend/access/transam/twophase.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/src/backend/access/transam/twophase.c b/src/backend/access/transam/twophase.c index e98286d768..db8d95f8cb 100644 --- a/src/backend/access/transam/twophase.c +++ b/src/backend/access/transam/twophase.c @@ -1505,6 +1505,7 @@ FinishPreparedTransaction(const char *gid, bool isCommit) GlobalTransaction gxact; PGPROC *proc; TransactionId xid; + bool ondisk; char *buf; char *bufptr; TwoPhaseFileHeader *hdr; @@ -1657,6 +1658,8 @@ FinishPreparedTransaction(const char *gid, bool isCommit) PredicateLockTwoPhaseFinish(xid, isCommit); + ondisk = gxact->ondisk; + /* Clear shared memory state */ RemoveGXact(gxact); @@ -1672,7 +1675,7 @@ FinishPreparedTransaction(const char *gid, bool isCommit) /* * And now we can clean up any files we may have left. */ - if (gxact->ondisk) + if (ondisk) RemoveTwoPhaseFile(xid, true); MyLockedGxact = NULL; -- 2.39.2 (Apple Git-143)