From: | Michael Paquier <michael(at)paquier(dot)xyz> |
---|---|
To: | Masahiko Sawada <sawada(dot)mshk(at)gmail(dot)com> |
Cc: | PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: Make pg_waldump report replication origin ID, LSN, and timestamp. |
Date: | 2021-12-08 07:31:42 |
Message-ID: | YbBfXpo8/O4B7dto@paquier.xyz |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On Mon, Dec 06, 2021 at 11:24:09PM +0900, Masahiko Sawada wrote:
> On Mon, Dec 6, 2021 at 5:09 PM Michael Paquier <michael(at)paquier(dot)xyz> wrote:
>> Shouldn't you check for parsed.origin_lsn instead? The replication
>> origin is stored there as far as I read EndPrepare().
>
> Also, looking at PrepareRedoAdd(), we check the replication origin id.
> So I think that it'd be better to check origin_id for consistency.
Okay, this consistency would make sense, then. Perhaps some comments
should be added to tell that?
>> Commit records check after XACT_XINFO_HAS_ORIGIN, but
>> xact_desc_abort() may include this information for ROLLBACK PREPARED
>> transactions so we could use the same logic as xact_desc_commit() for
>> the abort case, no?
>
> Good catch! I'll submit an updated patch.
Thanks!
--
Michael
From | Date | Subject | |
---|---|---|---|
Next Message | Kyotaro Horiguchi | 2021-12-08 07:35:15 | Re: Is it correct to update db state in control file as "shutting down" during end-of-recovery checkpoint? |
Previous Message | Michael Paquier | 2021-12-08 07:21:55 | Re: add recovery, backup, archive, streaming etc. activity messages to server logs along with ps display |