Re: BUG #19109: catalog lookup with the wrong snapshot during logical decoding causes coredump - Mailing list pgsql-bugs

From Haiyang Li
Subject Re: BUG #19109: catalog lookup with the wrong snapshot during logical decoding causes coredump
Date
Msg-id 3bb4539e-c342-4e15-94f1-803203f1072e.mohen.lhy@alibaba-inc.com
Whole thread Raw
In response to Re: BUG #19109: catalog lookup with the wrong snapshot during logical decoding causes coredump  (Michael Paquier <michael@paquier.xyz>)
List pgsql-bugs
From:Michael Paquier <michael@paquier.xyz> 2025-11-11日 10:12 wrote:

> That's unfortunate.  Having a reproducible test case that works with
>  upstream would speed up the analysis of the problem a lot.

Indeed. I can reproduce the core in original instance by using pg_logical_slot_peek_changes
fucntion. But I can’t reproduce in other new instances.  I will try to reproduce this issue again later.

While debugging the original instance using gdb, I found that txn 10722 was skipped due to
SnapBuildCurrentState(builder) < SNAPBUILD_FULL_SNAPSHOT, causing the catalog snapshot
to miss its commit. This suggests that the root cause may be in the new transaction handling
logic during snapshot building.

Regards,
Haiyang Li

pgsql-bugs by date:

Previous
From: Michael Paquier
Date:
Subject: Re: BUG #19109: catalog lookup with the wrong snapshot during logical decoding causes coredump
Next
From: Thomas Munro
Date:
Subject: Re: [EXTERNAL]Re: BUG #19094: select statement on postgres 17 vs postgres 18 is returning different/duplicate results