Re: Fix LOCK_TIMEOUT handling in slotsync worker

On Tue, Dec 9, 2025 at 11:50 AM Chao Li <li.evan.chao@gmail.com> wrote:
>
> > On Dec 9, 2025, at 14:12, Amit Kapila <amit.kapila16@gmail.com> wrote:
> >
> > On Tue, Dec 9, 2025 at 11:23 AM Chao Li <li.evan.chao@gmail.com> wrote:
> >>
> >>
> >> Yeah, I just searched and see similar messages:
> >>
> >> ```
> >> logical replication parallel apply worker for subscription \"%s\" will stop because the subscription owner's superuser privileges have been revoked
> >>
> >> logical replication worker for subscription \"%s\" will restart because the subscription owner's superuser privileges have been revoked
> >> ```
> >>
> >> I think the new phrase is better. Maybe “is triggered” could be “has been triggered”?
> >>
> >
> > My AI tool says:
> >
> > Both options are grammatically correct, but the nuance differs:
> > "will stop because promotion is triggered"
> > This uses the present tense ("is triggered"), which suggests the
> > promotion event is happening right now, concurrently with the stopping
> > action.
> > "will stop because promotion has been triggered"
> > This uses the present perfect tense ("has been triggered"), which
> > implies the promotion event already occurred and is the reason for the
> > upcoming stop.
> >
> > In this case, because ShutDownSlotSync() will wait for the slotsync
> > worker to exit, so the first one ("will stop because promotion is
> > triggered") fits better.
> >
>
> Make sense. Then Zhijie’s v2 looks good to me.
>

BTW, by mistake, I ended up pushing 0001 which I think in itself is
not a bad idea. However, we can improve it at least in HEAD as part of
patch[1] where we are making changes in the same part of code. Do you
think that is okay?

[1] - https://www.postgresql.org/message-id/CAFPTHDYHjqq53f1Cbata2MrV2nRBDe6XgxXfqv4tw4rcT2-Y8Q%40mail.gmail.com

-- 
With Regards,
Amit Kapila.