Horton

Re: Something is wrong with wal_compression

Michael Paquier <michael@paquier.xyz>

View thread View on postgr.es
From: Michael Paquier <michael@paquier.xyz>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Thomas Munro <thomas.munro@gmail.com>, Andrey Borodin <amborodin86@gmail.com>, Andres Freund <andres@anarazel.de>, Laurenz Albe <laurenz.albe@cybertec.at>, Justin Pryzby <pryzby@telsasoft.com>, pgsql-hackers@lists.postgresql.org
Date: 2023-01-31T03:50:47Z
Lists: pgsql-hackers 
On Mon, Jan 30, 2023 at 02:57:13PM +0900, Michael Paquier wrote:
> 1) means more test cycles, and perhaps we could enforce compression of
> WAL while on it?  At the end, my vote would just go for 3) and drop
> the whole scenario, though there may be an argument in 1).

And actually I was under the impression that 1) is not completely
stable either in the test because we rely on the return result of
txid_current() with IPC::Run::start, so a checkpoint forcing a flush
may not be able to do its work.  In order to bring all my animals back
to green, I have removed the test.
--
Michael

Commits

  1. Remove recovery test 011_crash_recovery.pl

  2. Add a txid_status function.