Horton

Re: Race condition in InvalidateObsoleteReplicationSlots()

Tom Lane <tgl@sss.pgh.pa.us>

View thread View on postgr.es
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Álvaro Herrera <alvherre@alvh.no-ip.org>
Cc: Andres Freund <andres@anarazel.de>, pgsql-hackers@postgresql.org, Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp>, Fujii Masao <fujii@postgresql.org>
Date: 2021-06-19T19:16:50Z
Lists: pgsql-hackers 
=?utf-8?Q?=C3=81lvaro?= Herrera <alvherre@alvh.no-ip.org> writes:
> It occurred to me that this could be made better by sigstopping both
> walreceiver and walsender, then letting only the latter run; AFAICS this
> makes the test stable.  I'll register this on the upcoming commitfest to
> let cfbot run it, and if it looks good there I'll get it pushed to
> master.  If there's any problem I'll just remove it before beta2 is
> stamped.

Hmm ... desmoxytes has failed this test once, out of four runs since
it went in:

https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=desmoxytes&dt=2021-06-19%2003%3A06%3A04

None of the other animals that have reported in so far are unhappy.
Still, maybe that's not a track record we want to have for beta2?

I've just launched a run on gaur, which given its dinosaur status
might be the most likely animal to have an actual portability problem
with this test technique.  If you want to wait a few hours to see what
it says, that'd be fine with me.

			regards, tom lane

Commits

  1. Add test case for obsoleting slot with active walsender, take 2

  2. Add test case for obsoleting slot with active walsender

  3. Return ReplicationSlotAcquire API to its original form

  4. Fix race condition in invalidating obsolete replication slots

  5. Fix issues in invalidation of obsolete replication slots.

  6. Allow users to limit storage reserved by replication slots