Re: BUG #17928: Standby fails to decode WAL on termination of primary

Поиск
Список
Период
Сортировка
От Alexander Lakhin
Тема Re: BUG #17928: Standby fails to decode WAL on termination of primary
Дата
Msg-id ee604ee6-f464-1502-0289-f1046c85c182@gmail.com
обсуждение исходный текст
Ответ на Re: BUG #17928: Standby fails to decode WAL on termination of primary  (Thomas Munro <thomas.munro@gmail.com>)
Ответы Re: BUG #17928: Standby fails to decode WAL on termination of primary  (Thomas Munro <thomas.munro@gmail.com>)
Список pgsql-bugs
19.09.2023 13:09, Thomas Munro wrote:
On Tue, Sep 19, 2023 at 9:53 PM Michael Paquier <michael@paquier.xyz> wrote:
On Tue, Sep 19, 2023 at 09:25:34PM +1200, Thomas Munro wrote:
The problem was that the Perl functions advance_XXX() would sometimes
see an "unflushed" insert position caused by hint records and other
incidental stuff, and then we'd write our contrived header at a
disconnected location after a bunch of (unwritten) zeroes.  The
solution in this version is to start each test with
emit_message($node, 0), which flushes a small record.


Thanks for the fixes!

I've tested this patch version on Debian 9 and it works fine now.
But as to wal_log_hints, I see that v7 patches work for REL_12_STABLE and
REL_13_STABLE, but not for other branches, though v6 patches definitely
failed on REL_12_STABLE and REL_13_STABLE too. So the behavior improved,
but not for all versions, for some reason (I haven't looked deeper yet).
Also I've noticed a few minor discrepancies in the patches.
A commit message in patch.15 contains the sentence "We should fix ...",
not "We could fix ...", as in commit messages for other versions.
Also, patch.12 contains a duplicated "Reviewed-by" string.
And may be "xlp_prev" is a typo?

Best regards,
Alexander

В списке pgsql-bugs по дате отправления:

Предыдущее
От: Thomas Munro
Дата:
Сообщение: Re: BUG #17928: Standby fails to decode WAL on termination of primary
Следующее
От: mah0x211
Дата:
Сообщение: Re: BUG #18096: In edge-triggered epoll and kqueue, PQconsumeInput/PQisBusy are insufficient for correct async ops.