Horton

Re: removing global variable ThisTimeLineID

Drouvot, Bertrand <bdrouvot@amazon.com>

View thread View on postgr.es
From: "Drouvot, Bertrand" <bdrouvot@amazon.com>
To: Robert Haas <robertmhaas@gmail.com>, Michael Paquier <michael@paquier.xyz>
Cc: PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2021-11-23T09:36:16Z
Lists: pgsql-hackers 
Hi,

On 11/8/21 6:49 PM, Robert Haas wrote:
> On Mon, Nov 8, 2021 at 8:27 AM Robert Haas <robertmhaas@gmail.com> wrote:
> Well I went through and it seems to be OK: all the existing callers of
> that function first verify that we're not in recovery. The patch to
> make logical decoding work in standby mode might change that,

Indeed, I think the logical decoding on standby patch just needs to 
change the Assert in GetWALInsertionTimeLine() to check 
SharedRecoveryState is RECOVERY_STATE_DONE or RECOVERY_STATE_ARCHIVE.

Would that make sense from your point of view to add the check on 
RECOVERY_STATE_ARCHIVE or do you think it would need to be more 
"sophisticated"?

Thanks

Bertrand

Commits

  1. More cleanup of 'ThisTimeLineID'.

  2. Change ThisTimeLineID from a global variable to a local variable.

  3. Remove all use of ThisTimeLineID global variable outside of xlog.c

  4. Don't set ThisTimeLineID when there's no reason to do so.

  5. When fetching WAL for a basebackup, report errors with a sensible TLI.

  6. Remove useless code from CreateReplicationSlot.