Horton

Re: lastOverflowedXid does not handle transaction ID wraparound

Stan Hu <stanhu@gmail.com>

View thread View on postgr.es
From: Stan Hu <stanhu@gmail.com>
To: Kyotaro Horiguchi <horikyota.ntt@gmail.com>
Cc: Andrey Borodin <x4mmm@yandex-team.ru>, Dmitry Dolgov <9erthalion6@gmail.com>, Pg Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2021-10-21T14:20:57Z
Lists: pgsql-hackers 
On Wed, Oct 20, 2021 at 9:01 PM Kyotaro Horiguchi
<horikyota.ntt@gmail.com> wrote:
>
> lastOverflowedXid is the smallest subxid that possibly exists but
> possiblly not known to the standby. So if all top-level transactions
> older than lastOverflowedXid end, that means that all the
> subtransactions in doubt are known to have been ended.

Thanks for the patch! I verified that it appears to reset
lastOverflowedXid properly.

I may not be understanding
https://github.com/postgres/postgres/blob/dc899146dbf0e1d23fb24155a5155826ddce34c9/src/backend/storage/ipc/procarray.c#L1326-L1327
correctly, but isn't lastOverflowedXid the last subxid for a given
top-level XID, so isn't it actually the largest subxid that possibly
exists?

Commits

  1. Reset lastOverflowedXid on standby when needed

  2. Fix variable shadowing in procarray.c.

  3. Move InRecovery and standbyState global vars to xlogutils.c.