From: Tom Lane <tgl@sss.pgh.pa.us>
To: Thomas Munro <thomas.munro@gmail.com>
cc: Peter Geoghegan <pg@bowt.ie>, Andres Freund <andres@anarazel.de>,
        Masahiko Sawada <sawada.mshk@gmail.com>,
        Andrew Dunstan <andrew@dunslane.net>,
        Michael Paquier <michael@paquier.xyz>,
        Anastasia Lubennikova <lubennikovaav@gmail.com>,
        pgsql-hackers <pgsql-hackers@postgresql.org>
Subject: Re: A test for replay of regression tests
In-reply-to: 
 <CA+hUKGLV3wzmYFbNs+TZ1+w0e=hc61fcvrF3OmutuaPBuZMd0w@mail.gmail.com>
References: 
 <CA+hUKG+Bncrfh0piJ83HUZ4cG+58TikGXwDux4sj7jvoMc=eNg@mail.gmail.com>
 <CA+hUKGK-q9utvxorA8sCwF3S4SzBmxxDgneP4rLqeHWdZxM4Gg@mail.gmail.com>
 <CA+hUKGLsrWbiCcMxBLRBAP6Z=ykFRHWzdmP9YKujSKoSnEJECQ@mail.gmail.com>
 <b10198e2-7411-7946-dfaa-3c8c5525d5c9@dunslane.net>
 <CA+hUKGKMd24n_Uau_HfKRhzARLgOm-_BbvqK4S6SY3Q=Kq=MCA@mail.gmail.com>
 <CAD21AoCNoWjYkdEtr+VDoF9v__V905AedKZ9iF=ArgCtrbxZqw@mail.gmail.com>
 <CA+hUKGKRtpz3qAg7=Yne_oJ8ZdeiT3JbPPCayjgHFgrKUeKViQ@mail.gmail.com>
 <CAH2-Wzk3RYzqdaOs5PUqoUjd=_LbWc7zdDzV6rR6D3kM3C=fNg@mail.gmail.com>
 <CA+hUKGJm_uex8rDfTcP6ig_nNTySD9DOKq9iyghTtuE0dg+t7g@mail.gmail.com>
 <CAH2-Wzm_mXH2mxn150tPmh1_kjQ9eC2dKdEr=UtHVOjnS2x3Sg@mail.gmail.com>
 <20220325041607.3vcnw34wxoucb5zy@alap3.anarazel.de>
 <CAH2-WznxZDjHaxYk2u_wpJysQGBLiVxCa550Ja7xxHm-+eTVNw@mail.gmail.com>
 <CA+hUKGLV3wzmYFbNs+TZ1+w0e=hc61fcvrF3OmutuaPBuZMd0w@mail.gmail.com>
Comments: In-reply-to Thomas Munro <thomas.munro@gmail.com>
	message dated "Mon, 28 Mar 2022 16:11:24 +1300"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <1258184.1648876239.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Sat, 02 Apr 2022 01:10:39 -0400
Message-ID: <1258185.1648876239@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/1258185.1648876239%40sss.pgh.pa.us>
Precedence: bulk

Thomas Munro <thomas.munro@gmail.com> writes:
> cfbot found another source of nondeterminism in the regression tests,
> due to the smaller shared_buffers used in this TAP test:

This failure seems related but not identical:

https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=3Dmyna&dt=3D2022-0=
4-02%2004%3A00%3A26

portals.out is expecting that the "foo25ns" cursor will read
starting at the beginning of tenk1, but it's starting somewhere
else, which presumably is a syncscan effect.

I think the fundamental instability here is that this TAP test is
setting shared_buffers small enough to allow the syncscan logic
to kick in where it does not in normal testing.  Maybe we should
just disable syncscan in this test script?

			regards, tom lane