Re: BGWORKER_INTERRUPTIBLE workers are never restarted after DROP/ALTER/CREATE DATABASE

On Wed, Nov 11, 2026 at 3:24 AM Zsolt Parragi <zsolt(dot)parragi(at)percona(dot)com> wrote:
>
> On Tue, 10 Nov 2026, Michael Paquier <michael(at)paquier(dot)xyz> wrote:
> > This last sentence should be gone. There is no point in sticking
> > comments for things that are non-conditions to begin with. This was
> > clearly AI-generated, as the kind of things that agents like adding
> > for nothing based on solely the input provided..
>
> Eh, sorry about that, looks like I managed to skim over this comment
> without reading it properly. Removed.
>
> On Tue, 10 Nov 2026, Iwata, Aya <iwata(dot)aya(at)jp(dot)fujitsu(dot)com> wrote:
> > “A signal is sent directly to the worker so that it can be restarted according to the worker's bgw_restart_time setting.”
>
> I also edited that part.
>
> v5 is only these two comment changes compared to v3.

Thanks. LGTM except from one nit in the added line:

"A signal is sent directly to the worker so that it can be restarted
according to the worker's bgw_restart_time setting"

"can be restarted based on its bgw_restart_time and exited status"
seems to help to clarify the rule like the doc.

--
Regards,
Xuneng Zhou
HighGo Software Co., Ltd.