Re: [PROPOSAL] Termination of Background Workers for ALTER/DROP DATABASE - Mailing list pgsql-hackers

From Michael Paquier
Subject Re: [PROPOSAL] Termination of Background Workers for ALTER/DROP DATABASE
Date
Msg-id aPBoXXW3XuwiIsHG@paquier.xyz
Whole thread Raw
In response to RE: [PROPOSAL] Termination of Background Workers for ALTER/DROP DATABASE  ("Hayato Kuroda (Fujitsu)" <kuroda.hayato@fujitsu.com>)
List pgsql-hackers
On Tue, Oct 14, 2025 at 04:36:01AM +0000, Hayato Kuroda (Fujitsu) wrote:
>> The only question I have is whether to
>> make sending a termination signal a default behaviour and let the flag
>> deactivate it.
>
> I think it can be done in future enhancement.
> We can change the default behavior based on the feedback from developers.

Yeah, I cannot agree with a change in the default behavior to cancel
the workers on a database touched by a database command.  This is a
behavior that exists since bgworkers are supported in tree in 9.3.  If
one is interested in making the workers more responsive, they could
just flip the flag switch.
--
Michael

Attachment

pgsql-hackers by date:

Previous
From: Fujii Masao
Date:
Subject: Re: Extend documentation for pg_stat_replication.backend_xmin
Next
From: Michael Paquier
Date:
Subject: Re: [PROPOSAL] Termination of Background Workers for ALTER/DROP DATABASE