From: "Alvaro Herrera" <alvherre@2ndquadrant.com>
> MauMau escribió:
>
>> One concern is that umount would fail in such a situation because
>> postgres has some open files on the filesystem, which is on the
>> shared disk in case of traditional HA cluster.
>
> See my reply to Noah. If postmaster stays around, would this be any
> different? I don't think so.
I'll consider this a bit and respond as a reply to Andres-san's mail.
> Actually, in further testing I noticed that the fast-path you introduced
> in BackendCleanup (or was it HandleChildCrash?) in the immediate
> shutdown case caused postmaster to fail to clean up properly after
> sending the SIGKILL signal, so I had to remove that as well. Was there
> any reason for that?
You are talking about the code below at the beginning of HandleChildCrash(),
aren't you?
/* Do nothing if the child terminated due to immediate shutdown */if (Shutdown == ImmediateShutdown) return;
If my memory is correct, this was necessary; without this,
HandleChildCrash() or LogChildExit() left extra messages in the server log.
LOG: %s (PID %d) exited with exit code %d
LOG: terminating any other active server processes
These messages are output because postmaster sends SIGQUIT to its children
and wait for them to terminate. The children exit with non-zero status when
they receive SIGQUIT.
Regards
MauMau