Horton

Re: XLogReadRecord() error in XlogReadTwoPhaseData()

Michael Paquier <michael@paquier.xyz>

View thread View on postgr.es
From: Michael Paquier <michael@paquier.xyz>
To: Noah Misch <noah@leadboat.com>
Cc: pgsql-hackers@postgresql.org
Date: 2021-11-08T04:42:46Z
Lists: pgsql-hackers 
On Sat, Nov 06, 2021 at 06:31:57PM -0700, Noah Misch wrote:
> As a first step, let's report the actual XLogReadRecord() error message.
> Attached.

Good catch!  This looks good.

> All the other sites that expect no error already do this.

Indeed.  Looking closer, I think that we'd better improve
DecodingContextFindStartpoint(),
pg_logical_replication_slot_advance(), XLogSendLogical() as well as
pg_logical_slot_get_changes_guts() to follow a format closer to what
you have in your patch, with an error message that describes the
context where the problem has been found, instead of just elog()'ing
what XLogReadRecord() returns.
--
Michael

Commits

  1. Under has_wal_read_bug, skip recovery/t/032_relfilenode_reuse.pl.

  2. Under has_wal_read_bug, skip contrib/bloom/t/001_wal.pl.

  3. Use Test::Builder::todo_start(), replacing $::TODO.

  4. On sparc64+ext4, suppress test failures from known WAL read failure.

  5. Report any XLogReadRecord() error in XlogReadTwoPhaseData().