Re: VM corruption on standby

On Wed, Aug 20, 2025 at 2:57 AM Andres Freund <andres@anarazel.de> wrote:
> On 2025-08-20 02:54:09 +1200, Thomas Munro wrote:
> > > On linux - the primary OS with OOM killer troubles - I'm pretty sure'll lwlock
> > > waiters would get killed due to the postmaster death signal we've configured
> > > (c.f. PostmasterDeathSignalInit()).
> >
> > No, that has a handler that just sets a global variable.  That was
> > done because recovery used to try to read() from the postmaster pipe
> > after replaying every record.  Also we currently have some places that
> > don't want to be summarily killed (off the top of my head, syncrep
> > wants to send a special error message, and the logger wants to survive
> > longer than everyone else to catch as much output as possible, things
> > I've been thinking about in the context of threads).
>
> That makes no sense. We should just _exit(). If postmaster has been killed,
> trying to stay up longer just makes everything more fragile. Waiting for the
> logger is *exactly* what we should *not* do - what if the logger also crashed?
> There's no postmaster around to start it.

Nobody is waiting for the logger.  The logger waits for everyone else
to exit first to collect forensics:

     * Unlike all other postmaster child processes, we'll ignore postmaster
     * death because we want to collect final log output from all backends and
     * then exit last.  We'll do that by running until we see EOF on the
     * syslog pipe, which implies that all other backends have exited
     * (including the postmaster).

The syncrep case is a bit weirder: it wants to tell the user that
syncrep is broken, so its own WaitEventSetWait() has
WL_POSTMASTER_DEATH, but that's basically bogus because the backend
can reach WaitEventSetWait(WL_EXIT_ON_PM_DEATH) in many other code
paths.  I've proposed nuking that before.