Horton

Re: Race condition in recovery?

Robert Haas <robertmhaas@gmail.com>

View thread View on postgr.es
From: Robert Haas <robertmhaas@gmail.com>
To: Dilip Kumar <dilipbalaut@gmail.com>
Cc: Kyotaro Horiguchi <horikyota.ntt@gmail.com>, hlinnaka <hlinnaka@iki.fi>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2021-06-09T16:38:36Z
Lists: pgsql-hackers 
On Wed, Jun 9, 2021 at 4:07 AM Dilip Kumar <dilipbalaut@gmail.com> wrote:
> Reason for the problem was that the "-Xnone" parameter was not
> accepted by "sub backup" in PostgresNode.pm so I created that for
> backpatch.  With attached patches I am to make it pass in v12,v11,v10
> (with fix) and fail (without fix).  However, we will have to make some
> change for 9.6 because pg_basebackup doesn't support -Xnone on 9.6,
> maybe we can delete the content from pg_wal after the backup, if we
> think that approach looks fine then I will make the changes for 9.6 as
> well.

Ah. I looked into this and found that this is because commit
081876d75ea15c3bd2ee5ba64a794fd8ea46d794 is new in master, so actually
that change is absent in all the back-branches. I have now back-ported
that portion of that commit to v13, v12, v11, and v10.

-- 
Robert Haas
EDB: http://www.enterprisedb.com

Commits

  1. Further refinement of stuck_on_old_timeline recovery test

  2. Adjust new test case to set wal_keep_segments.

  3. Adjust new test case to set wal_keep_size.

  4. Fix corner case failure of new standby to follow new primary.

  5. Back-port a few PostgresNode.pm methods.

  6. Allow PostgresNode.pm's backup method to accept backup_options.

  7. Add end-to-end testing of pg_basebackup's tar-format output.

  8. Allow PostgresNode.pm tests to wait for catchup

  9. Delay reading timeline history file until it's fetched from master.