Re: "make check" changes have caused buildfarm deterioration.

Michael Paquier wrote:
> On Tue, Jul 21, 2015 at 2:39 PM, Michael Paquier
> <michael(dot)paquier(at)gmail(dot)com> wrote:

> > Regarding initdb.log and postmaster.log, this is definitely a bug.
> > Those have been moved by dcae5fa from log/ to tmp_check/log/,
> > tmp_check/ getting removed at the end of pg_regress if there are no
> > failures counted. Both files will be saved in log/ at the location
> > pg_regress is called using outputdir whose default is ".". This way
> > behavior is similar to ~9.4. Attached is a patch to fix this for 9.5
> > and master.
>
> Something I just noticed: an entry for log/ in test_ddl_deparse's
> gitignore is missing.

Well, three things: 1) the entry is not missing right now, but it will
be missing if we apply your patch; 2) the file is inconsistent with the
other test modules anyway so we might as well apply your patch to make
them all alike; 3) we shouldn't really do anything about that until the
other patch's fate is decided.

So, regarding the other patch, I don't know if it's that useful to keep
the log files when the check succeeds; and if it fails, the only problem
we have, I think, is that the path is wrong in the buildfarm code and
that seems easily fixed, but do we want to make the master branch
different from the others? Maybe the BF code can look up the new path
first, and if it can't find the file then look in the old path.

Unless I'm misunderstanding the whole thing.

--
Álvaro Herrera http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services