Re: Reducing power consumption on idle servers

On Sun, 20 Nov 2022 at 20:00, Simon Riggs <simon(dot)riggs(at)enterprisedb(dot)com> wrote:
>
> On Thu, 24 Mar 2022 at 16:21, Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
> >
> > On Thu, Mar 24, 2022 at 12:02 PM Simon Riggs
>
> > > What changes will be acceptable for bgwriter, walwriter and logical worker?
> >
> > Hmm, I think it would be fine to introduce some kind of hibernation
> > mechanism for logical workers. bgwriter and wal writer already have a
> > hibernation mechanism, so I'm not sure what your concern is there
> > exactly. In your initial email you said you weren't proposing changes
> > there, but maybe that changed somewhere in the course of the
> > subsequent discussion. If you could catch me up on your present
> > thinking that would be helpful.
>
> Now that we seem to have solved the problem for Startup process, let's
> circle back to the others....

Thanks for committing changes to the startup process.

> Bgwriter does hibernate currently, but only at 50x the bgwriter_delay.
> At default values that is 5s, but could be much less. So we need to
> move that up to 60s, same as others.
>
> WALwriter also hibernates currently, but only at 25x the
> wal_writer_delay. At default values that is 2.5s, but could be much
> less. So we need to move that up to 60s, same as others. At the same
> time, make sure that when we hibernate we use a new WaitEvent,
> similarly to the way bgwriter reports its hibernation state (which
> also helps test the patch).

Re-attaching patch for bgwriter and walwriter, so it is clear this is
not yet committed.

I've added this to the next CF, since the entry was closed when the
startup patch was committed.

--
Simon Riggs http://www.EnterpriseDB.com/