Re: Performance degradation in commit ac1d794
| От | Amit Kapila |
|---|---|
| Тема | Re: Performance degradation in commit ac1d794 |
| Дата | |
| Msg-id | CAA4eK1+n_6U6Y7ug2zgE5VCzWQwQiJK8_nK1vPd8UnYWQaHaEw@mail.gmail.com обсуждение исходный текст |
| Ответ на | Re: Performance degradation in commit ac1d794 (Andres Freund <andres@anarazel.de>) |
| Список | pgsql-hackers |
On Sat, Dec 26, 2015 at 5:41 PM, Andres Freund <andres@anarazel.de> wrote:
On 2015-12-26 12:22:48 +0100, Andres Freund wrote:
> > > 3) Replace the postmaster_alive_fds socketpair by some other signalling
> > > mechanism. E.g. sending a procsignal to each backend, which sets the
> > > latch and a special flag in the latch structure.
> >
> > And what would send the signal? The entire point here is to notice the
> > situation where the postmaster has crashed. It can *not* depend on the
> > postmaster taking some action.
>
> Ahem. Um. Look, over there --->
>
> I blame it on all the food.
A unportable and easy version of this, actually making sense this time,
would be to use prctl(PR_SET_PDEATHSIG, SIGQUIT). That'd send SIGQUIT to
backends whenever postmaster dies. Obviously that's not portable
either - doing this for linux only wouldn't be all that kludgey tho.
There is a way to make backends exit in Windows as well by using
JobObjects and use limitFlags as JOB_OBJECT_LIMIT_KILL_ON_JOB_CLOSE
for JOBOBJECT_BASIC_LIMIT_INFORMATION.
В списке pgsql-hackers по дате отправления: