Re: Logical replication keepalive flood

On Thu, Sep 30, 2021 at 1:26 PM Kyotaro Horiguchi
<horikyota(dot)ntt(at)gmail(dot)com> wrote:
>
> At Thu, 30 Sep 2021 16:21:25 +1000, Greg Nancarrow <gregn4422(at)gmail(dot)com> wrote in
> > On Thu, Sep 30, 2021 at 3:56 PM Amit Kapila <amit(dot)kapila16(at)gmail(dot)com> wrote:
> > >
>
> > > Also, more to the point this special keep_alive seems to be sent for
> > > synchronous replication and walsender shutdown as they can expect
> > > updated locations. You haven't given any reason (theory) why those two
> > > won't be impacted due to this change? I am aware that for synchronous
> > > replication, we wake waiters while ProcessStandbyReplyMessage but I am
> > > not sure how it helps with wal sender shutdown? I think we need to
> > > know the reasons for this message and then need to see if the change
> > > has any impact on the same.
> > >
> >
> > Yes, I'm not sure about the possible impacts, still looking at it.
>
> If the comment describes the objective correctly, the only possible
> impact would be that there may be a case where server responds a bit
> slowly for a shutdown request. But I'm not sure it is definitely
> true.
>

So, we should try to find how wal sender shutdown is dependent on
sending keep alive and second thing is what about sync rep case? I
think in the worst case that also might delay. Why do you think that
would be acceptable?

--
With Regards,
Amit Kapila.