Feedback-ID: id4a34324:Fastmail
Date: Tue, 31 Jan 2023 13:04:24 -0800
From: Andres Freund <andres@anarazel.de>
To: Thomas Munro <thomas.munro@gmail.com>,
 Justin Pryzby <pryzby@telsasoft.com>
CC: Michael Paquier <michael@paquier.xyz>, pgsql-hackers@lists.postgresql.org,
 Nazir Bilal Yavuz <byavuz81@gmail.com>
Subject: Re: pg_upgrade test failure
User-Agent: K-9 Mail for Android
In-Reply-To: 
 <CA+hUKGKCVy2=o=d8c2Va6a_3Rpf_KkhUitkWCZ3hzuO2VwLMXA@mail.gmail.com>
References: 
 <CA+hUKG+bk4KGwYT4JrZEdAgNdFm-c8ezTrxf-PacS-npQ_HdSw@mail.gmail.com>
 <YzqBIzMwoBiESmkk@paquier.xyz>
 <CA+hUKGKsdzw06c5nnb=KYG9GmvyykoVpJA_VR3k0r7cZOKcx6Q@mail.gmail.com>
 <Y04mN0ZLNzJywrad@paquier.xyz> <20221018043144.GF7745@telsasoft.com>
 <20221018162908.cb3n7tahtekypbkf@awork3.anarazel.de>
 <CA+hUKGKFh-F2fz_18j3N6bKr966gmaJxV40T0qhDwhKkW+291g@mail.gmail.com>
 <20221206181533.pmcz75zbjgvnnzum@awork3.anarazel.de>
 <CA+hUKGJTM5FswRV-_JWzVLdw=Lq1pjNsvGfONRJp=dpR6YR81Q@mail.gmail.com>
 <CA+hUKGLfruDhC_8QRGDoCndsdzDVjrf7TYwgsv+OfoAmK2syEg@mail.gmail.com>
 <20230131172806.GM22427@telsasoft.com>
 <CA+hUKGKCVy2=o=d8c2Va6a_3Rpf_KkhUitkWCZ3hzuO2VwLMXA@mail.gmail.com>
Message-ID: <F0554EC4-4545-4224-B373-9B01EA8D25F1@anarazel.de>
MIME-Version: 1.0
Content-Type: text/plain;
 charset=utf-8
Content-Transfer-Encoding: quoted-printable
Archived-At: 
 <https://www.postgresql.org/message-id/F0554EC4-4545-4224-B373-9B01EA8D25F1%40anarazel.de>
Precedence: bulk

Hi,=20

On January 31, 2023 12:54:42 PM PST, Thomas Munro <thomas=2Emunro@gmail=2E=
com> wrote:
>On Wed, Feb 1, 2023 at 6:28 AM Justin Pryzby <pryzby@telsasoft=2Ecom> wro=
te:
>> > I pushed the rmtree() change=2E  Let's see if that helps, or tells us
>> > something new=2E
>>
>> I found a few failures since then:
>>
>> https://api=2Ecirrus-ci=2Ecom/v1/artifact/task/6696942420361216/testrun=
/build/testrun/pg_upgrade/002_pg_upgrade/log/regress_log_002_pg_upgrade
>>
>> pg_upgrade: warning: could not remove directory "C:/cirrus/build/testru=
n/pg_upgrade/002_pg_upgrade/data/t_002_pg_upgrade_new_node_data/pgdata/pg_u=
pgrade_output=2Ed/20230131T134931=2E720/log": Directory not empty
>> pg_upgrade: warning: could not remove directory "C:/cirrus/build/testru=
n/pg_upgrade/002_pg_upgrade/data/t_002_pg_upgrade_new_node_data/pgdata/pg_u=
pgrade_output=2Ed/20230131T134931=2E720": Directory not empty
>
>So no change: we didn't see "could not unlink file =2E=2E=2E"=2E  So I th=
ink
>that means that it was rmtree() that unlinked the file for the *first*
>time, but someone else has it open=2E
>
>Even though Windows is at this point eroding my love of computers and
>making me consider a new career in, I dunno, carrot farming or
>something, I have one more idea=2E  Check out this kluge in
>src/bin/pg_upgrade/exec=2Ec:
>
>        /*
>         * "pg_ctl -w stop" might have reported that the server has stopp=
ed
>         * because the postmaster=2Epid file has been removed, but "pg_ct=
l -w
>         * start" might still be in the process of closing and might stil=
l be
>         * holding its stdout and -l log file descriptors open=2E  Theref=
ore,
>         * try to open the log file a few more times=2E
>         */
>
>I'm not sure about anything, but if that's what's happening here, then
>maybe the attached would help=2E  In short, it would make the previous
>theory true (the idea of a second unlink() saving the day)=2E


Maybe we should just handle it by sleeping and retrying, if on windows? Sa=
d to even propose=2E=2E=2E=20
--=20
Sent from my Android device with K-9 Mail=2E Please excuse my brevity=2E