Re: Setting pd_lower in GIN metapage

Alvaro Herrera <alvherre@2ndquadrant.com>

From: Alvaro Herrera <alvherre@2ndquadrant.com>
To: Masahiko Sawada <sawada.mshk@gmail.com>
Cc: Michael Paquier <michael.paquier@gmail.com>, Amit Langote <Langote_Amit_f8@lab.ntt.co.jp>, Amit Kapila <amit.kapila16@gmail.com>, Tom Lane <tgl@sss.pgh.pa.us>, Pg Hackers <pgsql-hackers@postgresql.org>
Date: 2017-06-23T16:35:40Z
Lists: pgsql-hackers
Masahiko Sawada wrote:
> On Fri, Jun 23, 2017 at 3:31 PM, Michael Paquier
> <michael.paquier@gmail.com> wrote:
> > On Fri, Jun 23, 2017 at 3:17 PM, Amit Langote
> > <Langote_Amit_f8@lab.ntt.co.jp> wrote:
> >> Initially, I had naively set wal_consistency_check = all before running
> >> make installcheck and then had to wait for a long time to confirm that WAL
> >> generated by the gin test indeed caused consistency check failure on the
> >> standby with the v1 patch.
> >
> > wal_consistency_check = gin would have saved you a lot of I/O.
> >
> >> But I can see Sawada-san's point that there should be some way for
> >> developers writing code that better had gone through WAL consistency
> >> checking facility to do it without much hassle.  But then again, it may
> >> not be that frequent to need that.
> 
> Yeah, it should be optional. I imagined providing such an option of
> pg_regress or TAP test for the developers.

As far as I know it is possible to have third-party modules that extend
the buildfarm client script so that it runs additional tests that the
standard ones.  You could have a custom module installed in some
powerful machine of yours that runs the WAL consistency check and report
the results to the buildfarm.  A single animal running that test should
be enough, right?

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


Commits

  1. Flag index metapages as standard-format in xlog.c calls.

  2. Set the metapage's pd_lower correctly in brin, gin, and spgist indexes.