Re: [PATCH] Add recovery boundary WAL record for database and tablespace commands

From: JoongHyuk Shin <sjh910805(at)gmail(dot)com>
To: João Rito <joaovictor(dot)rito(at)gmail(dot)com>
Cc: zhong950419(at)gmail(dot)com, pgsql-hackers(at)lists(dot)postgresql(dot)org
Subject: Re: [PATCH] Add recovery boundary WAL record for database and tablespace commands
Date: 2026-11-14 10:53:39
Message-ID: CACSdjfPq_GWudFkEo1ExCBNeNAzpsLdNa37j_hdXzY5ObymA5g@mail.gmail.com
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Hi João,

Thanks for testing v3.

On Fri, Nov 13, 2026 at 9:17 PM João Rito <joaovictor(dot)rito(at)gmail(dot)com> wrote:
> Maybe a short sentence in the
> docs could say that a target falling inside that window can still leave
> this state, so users pick a time before the command.

I'd rather not add it. The state you saw is what recovery leaves when
it stops before the commit of one of these commands, with the changes
the command makes before committing already replayed. A time target is
not the only way to end up there. With recovery_target_xid set to the
command's transaction and recovery_target_inclusive = off, recovery
stops at the same place. I kept the docs as they are for that case,
and I don't think the state should be described under
recovery_target_time alone. The text v3 adds there says where recovery
now stops, and I'd prefer to keep it to that.

--
JoongHyuk Shin

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Alexander Korotkov 2026-11-14 11:00:11 Re: test: avoid redundant standby catchup in 049_wait_for_lsn
Previous Message Robert I Nilsson 2026-11-14 10:41:14 Re: [PATCH] Fix SASLprep validation after NFKC normalization