Horton

Re: prevent immature WAL streaming

Alvaro Herrera <alvherre@alvh.no-ip.org>

View thread View on postgr.es
From: Alvaro Herrera <alvherre@alvh.no-ip.org>
To: Andrew Dunstan <andrew@dunslane.net>
Cc: Tom Lane <tgl@sss.pgh.pa.us>, "Bossart, Nathan" <bossartn@amazon.com>, Kyotaro Horiguchi <horikyota.ntt@gmail.com>, "andres@anarazel.de" <andres@anarazel.de>, "masao.fujii@oss.nttdata.com" <masao.fujii@oss.nttdata.com>, "pgsql-hackers@lists.postgresql.org" <pgsql-hackers@lists.postgresql.org>, "mengjuan.cmj@alibaba-inc.com" <mengjuan.cmj@alibaba-inc.com>, "Jakub.Wartak@tomtom.com" <Jakub.Wartak@tomtom.com>, Ryo Matsumura <matsumura.ryo@fujitsu.com>
Date: 2021-09-29T21:27:54Z
Lists: pgsql-hackers 
On 2021-Sep-29, Andrew Dunstan wrote:

> > The relevant info seems to be
> >
> > # Running: pg_basebackup -D /home/pgrunner/bf/root/REL_14_STABLE/pgsql.build/src/test/recovery/tmp_check/t_026_overwrite_contrecord_primary2_data/backup/backup -h 127.0.0.1 -p 59502 --checkpoint fast --no-sync
> > pg_basebackup: error: connection to server at "127.0.0.1", port 59502 failed: FATAL:  no pg_hba.conf entry for replication connection from host "127.0.0.1", user "pgrunner", no encryption
> > Bail out!  system pg_basebackup failed
> >
> > which looks like a pretty straightforward bogus-connection-configuration
> > problem, except why wouldn't other BF members show it?
> 
> This:
> 
>     # Second test: a standby that receives WAL via archive/restore commands.
>     $node = PostgresNode->new('primary2');
>     $node->init(
>         has_archiving => 1,
>         extra         => ['--wal-segsize=1']);
> 
> doesn't have "allows_streaming => 1".

Hmm, but I omitted allows_streaming on purpose -- I only wanted
archiving, not streaming.  I understand that your point is that
set_replication_conf is not called unless allows_streaming is set.

So, do we take the stance that we have no right to expect pg_basebackup
to work if we didn't pass allow_streaming => 1?  If so, the fix is to
add it.  But my preferred fix would be to call set_replication_conf if
either allows_streaming or has_archiving are given.


Another easy fix would be to call $primary2->set_replication_conf in the
test file, but then you'd complain that that's supposed to be an
internal method :-)

(This reminds me that I had to add something that seemed like it should
have been unnecessary: wal_level=replica should become set if I request
archiving, right?  Otherwise the WAL archive is useless.  I also had to
add max_wal_senders=2 so that pg_basebackup would work, but I'm on the
fence about setting that automatically if has_archiving is given.)

-- 
Álvaro Herrera              Valdivia, Chile  —  https://www.EnterpriseDB.com/
"At least to kernel hackers, who really are human, despite occasional
rumors to the contrary" (LWN.net)

Commits

  1. Fix determination of broken LSN in OVERWRITTEN_CONTRECORD

  2. Fix instability in 026_overwrite_contrecord.pl test.

  3. Fix typo

  4. Change recently added test code for stability

  5. Remove unstable, unnecessary test; fix typo

  6. Repair two portability oversights of new test

  7. Fix WAL replay in presence of an incomplete record

  8. Document XLOG_INCLUDE_XID a little better

  9. Revert "Avoid creating archive status ".ready" files too early"

  10. Avoid creating archive status ".ready" files too early

  11. Fix scenario where streaming standby gets stuck at a continuation record.

  12. Teach xlogreader to follow timeline switches

  13. Forbid using pg_xlogfile_name() and pg_xlogfile_name_offset() during