Re: Race condition in recovery?

Поиск
Список
Период
Сортировка
От Dilip Kumar
Тема Re: Race condition in recovery?
Дата
Msg-id CAFiTN-uaf7kaW32-gOJ3sa8eCnfyy_y_pDnDhKVFvBzr-uXkJA@mail.gmail.com
обсуждение исходный текст
Ответ на Re: Race condition in recovery?  (Robert Haas <robertmhaas@gmail.com>)
Ответы Re: Race condition in recovery?  (Robert Haas <robertmhaas@gmail.com>)
Список pgsql-hackers
On Wed, May 26, 2021 at 9:40 PM Robert Haas <robertmhaas@gmail.com> wrote:

> ...which has a clear race condition.
> src/test/recovery/t/023_pitr_prepared_xact.pl has logic to wait for a
> WAL file to be archived, so maybe we can steal that logic and use it
> here.

Yeah, done that, I think we can use exact same logic for history files
as well because if wal file is archived then history file must be
because a) history file get created during promotion so created before
WAL file with new TL is ready for archive b) Archiver archive history
files before archiving any WAL files.

src/test/recovery/t/025_stuck_on_old_timeline.pl

> I suggest we rename the test to something a bit more descriptive. Like
> instead of 025_timeline_issue.pl, perhaps
> 025_stuck_on_old_timeline.pl? Or I'm open to other suggestions, but
> "timeline issue" is a bit too vague for my taste.

Changed as suggested.

--
Regards,
Dilip Kumar
EnterpriseDB: http://www.enterprisedb.com

Вложения

В списке pgsql-hackers по дате отправления:

Предыдущее
От: Kyotaro Horiguchi
Дата:
Сообщение: Re: Bracket, brace, parenthesis
Следующее
От: Masahiko Sawada
Дата:
Сообщение: Re: Skipping logical replication transactions on subscriber side