Re: Addition of %b/backend_type in log_line_prefix of TAP test logs

From: Michael Paquier <michael(at)paquier(dot)xyz>
To: Daniel Gustafsson <daniel(at)yesql(dot)se>
Cc: Fujii Masao <masao(dot)fujii(at)gmail(dot)com>, Postgres hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>
Subject: Re: Addition of %b/backend_type in log_line_prefix of TAP test logs
Date: 2025-05-21 23:50:04
Message-ID: aC5mrMdy9St42vFQ@paquier.xyz
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Wed, May 21, 2025 at 09:24:26AM +0200, Daniel Gustafsson wrote:
> Agreed, and +1 on the change in general.

Another point perhaps worth considering after a second look: how about
pg_ci_base.conf? We have the following line there, exposing this
configuration with TEMP_CONFIG:
src/tools/ci/pg_ci_base.conf:log_line_prefix = '%m [%p][%b] %q[%a][%v:%x] '

The two other ones are:
src/test/perl/PostgreSQL/Test/Cluster.pm: '%m [%p] %q%a '
src/test/regress/pg_regress.c: '%m %b[%p] %q%a '

Perhaps it would be better to apply the same configuration as the CI
by default at the end? The difference is that the CI exposes
[virtual] XIDs. I've never used this information for debugging, but
that may be worth considering for consistency? This would make easier
checks across the buildfarm, the CI or local machines when looking for
differences in the logs with the same tests.

(I was incorrect previously about CI not using backend_type upthread.)
--
Michael

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Jeff Davis 2025-05-21 23:53:17 Re: Statistics Import and Export
Previous Message Tom Lane 2025-05-21 23:48:05 Re: [PATCH] Add pretty-printed XML output option