Horton

Re: WAL segments removed from primary despite the fact that logical replication slot needs it.

Amit Kapila <amit.kapila16@gmail.com>

View thread
From: Amit Kapila <amit.kapila16@gmail.com>
To: depesz@depesz.com
Cc: pgsql-bugs mailing list <pgsql-bugs@postgresql.org>
Date: 2022-11-14T16:22:05Z
Lists: pgsql-bugs

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Fix a possibility of logical replication slot's restart_lsn going backwards.

On Mon, Nov 14, 2022 at 7:03 PM hubert depesz lubaczewski
<depesz@depesz.com> wrote:
>
> On Mon, Nov 14, 2022 at 06:30:57PM +0530, Amit Kapila wrote:
> > > There is something weird happening:
> > What exactly weird you are seeing in this? To me, it appears as if the
> > system due to some reason ignores an existing slot that has
> > restart_lsn as 1039D/83825958.
>
> The weird part for me is that it is trying to remove wal files older
> than the same "x" many times.
>

I think that is okay because as per checkpointer's computation it
decides not to remove/replace any new WAL files. At this stage, I am
not getting any idea except for getting the value of
XLogGetReplicationSlotMinimumLSN() in one of the LOG prints. If you
can't add all the LOGs, I shared in the last patch, can you try to get
the value of XLogGetReplicationSlotMinimumLSN() by appending to the
existing LOG "attempting to remove WAL segments older than log file
.."?

-- 
With Regards,
Amit Kapila.