Horton

Re: v13: CLUSTER segv with wal_level=minimal and parallel index creation

Kyotaro Horiguchi <horikyota.ntt@gmail.com>

View thread View on postgr.es
From: Kyotaro Horiguchi <horikyota.ntt@gmail.com>
To: tgl@sss.pgh.pa.us
Cc: pryzby@telsasoft.com, pgsql-hackers@postgresql.org, noah@leadboat.com
Date: 2020-09-07T08:40:36Z
Lists: pgsql-hackers

Attachments

At Mon, 07 Sep 2020 13:45:28 +0900 (JST), Kyotaro Horiguchi <horikyota.ntt@gmail.com> wrote in 
> The cause is that the worker had received pending-sync entry correctly
> but not never created a relcache entry for the relation using
> RelationBuildDesc. So the rd_firstRelfilenodeSubid is not correctly
> set.
> 
> I'm investigating it.

Relcaches are loaded from a file with old content at parallel worker
startup. The relcache entry is corrected by invalidation at taking a
lock but pending syncs are not considered.

Since parallel workers don't access the files so we can just ignore
the assertion safely, but I want to rd_firstRelfilenodeSubid flag at
invalidation, as attached PoC patch.

regards.

-- 
Kyotaro Horiguchi
NTT Open Source Software Center

Commits

  1. Fix rd_firstRelfilenodeSubid for nailed relations, in parallel workers.

  2. Skip WAL for new relfilenodes, under wal_level=minimal.