Message-ID: <4a7b82d8-17ad-83ae-4aa1-563c3c64c4c3@gmail.com>
Date: Tue, 17 Sep 2024 09:00:00 +0300
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.4.2
Subject: Re: [HACKERS] make async slave to wait for lsn to be replayed
Content-Language: en-US
To: Alexander Korotkov <aekorotkov@gmail.com>
Cc: Kevin Hale Boyes <kcboyes@gmail.com>,
 Kartyshov Ivan <i.kartyshov@postgrespro.ru>,
 Kyotaro Horiguchi <horikyota.ntt@gmail.com>, hlinnaka@iki.fi,
 alvherre@alvh.no-ip.org, pashkin.elfe@gmail.com,
 bharath.rupireddyforpostgres@gmail.com, euler@eulerto.com,
 thomas.munro@gmail.com, peter@eisentraut.org, amit.kapila16@gmail.com,
 dilipbalaut@gmail.com, smithpb2250@gmail.com,
 pgsql-hackers@lists.postgresql.org
References: <ca85a3c56e3c26cc29d1df14021845f6@postgrespro.ru>
 <CAPpHfdvGRssjqwX1+idm5Tu-eWsTcx6DthB2LhGqA1tZ29jJaw@mail.gmail.com>
 <CAPpHfdvDjrM6=XxDiuZBa+bg2qL34jwafa3uXf0Zsxm2PKOChg@mail.gmail.com>
 <20240620.173045.1163445432733259320.horikyota.ntt@gmail.com>
 <8d0286f9f0c2f9fc29794dda6b5f4600@postgrespro.ru>
 <CAPpHfdvwMn21n-zWecvr5gEzFy7rmszwbOTekjTE4taCe9frBw@mail.gmail.com>
 <CAPpHfds1QW-zZ_Hs4dkvX2pji=Ms1VE2C=oGFRu2xFbDA_-MgA@mail.gmail.com>
 <CAPpHfdt8rs0=NC3aHgcj_ZXGKEexZtk=Y8ytK188Ta3NWD5+Pg@mail.gmail.com>
 <CADAecHWKpaPuPGXAMOH=wmhTpydHWGPOk9KWX97UYhp5GdqCWw@mail.gmail.com>
 <CAPpHfdu5QN+ZGACS+7foxmr8_nekgA2PA+-G3BuOUrdBLBFb6Q@mail.gmail.com>
 <CAPpHfdtRZ+3use8n3Vc9RFVitXj59koTKuYLM4+3M0_En7BY9g@mail.gmail.com>
 <CAPpHfdsyZFEu-M5YLeX68cBG_N4j7R4TM6ROpP6wnNB_oDNKZA@mail.gmail.com>
 <CAPpHfdsgKwR5ifuBTdB4sOA1=h28=0tUHjaWRWi0C_VZR5PMdg@mail.gmail.com>
 <1d7b08f2-64a2-77fb-c666-c9a74c68eeda@gmail.com>
 <CAPpHfdv=5X5k=gRu0Q2cJ7pyWa0fqdnDFCPHQqUcrp8eypQWRg@mail.gmail.com>
From: Alexander Lakhin <exclusion@gmail.com>
In-Reply-To: 
 <CAPpHfdv=5X5k=gRu0Q2cJ7pyWa0fqdnDFCPHQqUcrp8eypQWRg@mail.gmail.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Archived-At: 
 <https://www.postgresql.org/message-id/4a7b82d8-17ad-83ae-4aa1-563c3c64c4c3%40gmail.com>
Precedence: bulk

Hi Alexander,

16.09.2024 21:55, Alexander Korotkov wrote:
> Please find two patches attached.  The first one does minor cleanup
> including misuse of words you've pointed.  The second one adds missing
> wait_for_catchup().  That should fix the test failure you've spotted.
> Please, check if it fixes an issue for you.

Thank you for looking at that!

Unfortunately, the issue is still here — the test failed for me 6 out of
10 runs, as below:
[05:14:02.807](0.135s) ok 8 - got error after standby promote
error running SQL: 'psql:<stdin>:1: ERROR:  recovery is not in progress
HINT:  Waiting for LSN can only be executed during recovery.'
while running 'psql -XAtq -d port=12734 host=/tmp/04hQ75NuXf dbname='postgres' -f - -v ON_ERROR_STOP=1' with sql 'CALL 
pg_wal_replay_wait('0/300F248');' at .../src/test/recovery/../../../src/test/perl/PostgreSQL/Test/Cluster.pm line 2140.

043_wal_replay_wait_standby.log:
2024-09-17 05:14:02.714 UTC [1817258] 043_wal_replay_wait.pl ERROR: recovery is not in progress
2024-09-17 05:14:02.714 UTC [1817258] 043_wal_replay_wait.pl DETAIL:  Recovery ended before replaying target LSN 
2/570CD648; last replay LSN 0/300F210.
2024-09-17 05:14:02.714 UTC [1817258] 043_wal_replay_wait.pl STATEMENT:  CALL pg_wal_replay_wait('2/570CD648');
2024-09-17 05:14:02.714 UTC [1817155] LOG:  checkpoint starting: force
2024-09-17 05:14:02.714 UTC [1817154] LOG:  database system is ready to accept connections
2024-09-17 05:14:02.811 UTC [1817270] 043_wal_replay_wait.pl LOG: statement: CALL pg_wal_replay_wait('0/300F248');
2024-09-17 05:14:02.811 UTC [1817270] 043_wal_replay_wait.pl ERROR: recovery is not in progress

pg_waldump -p .../t_043_wal_replay_wait_primary_data/pgdata/pg_wal/ 000000010000000000000003
rmgr: Transaction len (rec/tot):     34/    34, tx:        748, lsn: 0/0300F1E8, prev 0/0300F1A8, desc: COMMIT 
2024-09-17 05:14:01.654874 UTC
rmgr: Standby     len (rec/tot):     50/    50, tx:          0, lsn: 0/0300F210, prev 0/0300F1E8, desc: RUNNING_XACTS 
nextXid 749 latestCompletedXid 748 oldestRunningXid 749

I wonder, can you reproduce this with that bgwriter's modification?

I've also found two more "achievements" coined by 3c5db1d6b:
doc/src/sgml/func.sgml:   It may also happen that target <acronym>lsn</acronym> is not achieved
src/backend/access/transam/xlog.c-       * recovery was ended before achieving the target LSN.

Best regards,
Alexander