Horton

Re: [HACKERS] make async slave to wait for lsn to be replayed

Ivan Kartyshov <i.kartyshov@postgrespro.ru>

View thread View on postgr.es
From: Kartyshov Ivan <i.kartyshov@postgrespro.ru>
To: Kyotaro Horiguchi <horikyota.ntt@gmail.com>
Cc: aekorotkov@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
Date: 2024-07-10T04:58:23Z
Lists: pgsql-hackers 
>> I think the race condition you mentioned refers to the inconsistency
>> between the inHeap flag and the pairing heap caused by a race
>> condition between timeout and wakeup (or perhaps other combinations?
>> I'm not sure which version of the patch the mentioned race condition
>> refers to). However, I imagine it is difficult to reliably reproduce
>> this condition. In that regard, in the latest patch, the coherence
>> between the inHeap flag and the pairing heap is protected by LWLock,
>> so I believe we no longer need that test.
> 
> No, Alexandre means that Heikki point on race condition just before
> LWLock. But injection point we can inject and stepin on backend, and
> WaitLSNSetLatches is used from Recovery process. But I have trouble
> to wakeup injection point on server.

One more thing. I want to point, when you move injection point to
contrib dir and do the same test (044_wal_replay_wait_injection_test.pl)
step by step with hands, wakeup works well.

-- 
Ivan Kartyshov
Postgres Professional: www.postgrespro.com

Commits

  1. Ensure standby promotion point in 043_wal_replay_wait.pl

  2. Minor cleanup related to pg_wal_replay_wait() procedure

  3. Adjust pg_wal_replay_wait() procedure behavior on promoted standby

  4. pg_wal_replay_wait(): Fix typo in the doc

  5. Implement pg_wal_replay_wait() stored procedure

  6. Revert: Implement pg_wal_replay_wait() stored procedure

  7. Call WaitLSNCleanup() in AbortTransaction()

  8. Clarify what is protected by WaitLSNLock

  9. Use an LWLock instead of a spinlock in waitlsn.c

  10. Use the pairing heap instead of a flat array for LSN replay waiters

  11. Minor improvements for waitlsn.c

  12. Make the order of the header file includes consistent