Re: Disallow setting client_min_messages > ERROR?

Поиск
Список
Период
Сортировка
От Jonah H. Harris
Тема Re: Disallow setting client_min_messages > ERROR?
Дата
Msg-id CADUqk8Usa8EqhG_nLEApFXyu1UqMOA5y2MJq2+y9MD17e4uoTA@mail.gmail.com
обсуждение исходный текст
Ответ на Re: Disallow setting client_min_messages > ERROR?  (Isaac Morland <isaac.morland@gmail.com>)
Список pgsql-hackers
On Tue, Nov 6, 2018 at 2:46 PM Isaac Morland <isaac.morland@gmail.com> wrote:
On Tue, 6 Nov 2018 at 14:07, Jonah H. Harris <jonah.harris@gmail.com> wrote:
Two options presented:

- Hard patch removes FATAL/PANIC from client_message_level_options in guc.c, which also seems to make sense in regard to it's double-usage with trace_recovery_messages.

- Soft patch keeps FATAL/PANIC in client_message_level_options but coerces client_min_messages to ERROR when set to FATAL/PANIC and issues a warning. This also exports error_severity from elog.c to retrieve severity -> text mappings for the warning message.

 
What about no-op (soft patch) for 11.1 and backpatches, error (hard patch) for 12?

I'm usually a fan of the hard fix... but I do see the point they've made about during an upgrade.

Also, fixed wording in the soft patch (frontend protocol requires %s or above -> frontend protocol requires %s or below) attached.

--
Jonah H. Harris

Вложения

В списке pgsql-hackers по дате отправления:

Предыдущее
От: Peter Eisentraut
Дата:
Сообщение: Re: [HACKERS] generated columns
Следующее
От: Ondřej Bouda
Дата:
Сообщение: Re: backend crash on DELETE, reproducible locally