httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Colm MacCarthaigh <>
Subject Re: fixing graceful-stop with event mpm
Date Tue, 21 Aug 2007 10:26:15 GMT
On Mon, Aug 20, 2007 at 04:00:48PM -0700, Paul Querna wrote:
> Short: We need to call ap_close_listeners() earlier or more aggressively.
> Question: Where/How?
> Looking at the Event MPM in both trunk and 2.2.x, the listener_thread is
> where we call ap_close_listeners().  This does not seem to be working
> quickly enough, per PR 43081:

Crap :( When I test it on my linux box, it works for me, it's pretty
fast, I'm guessing I have the wrong sort of load to see this, any idea
how to replicate?

> I haven't quite got my head around how the listener thread is apparently
> blocking for so long, but I can't figure out another way that the
> listeners would stay online for so long either.

The listener thread isn't neccessarily the only thread with a copy of
the fd, I remember cgid catching me out before (which I fixed iirc). 

> Solution: Add call to ap_close_listeners() in wakeup_listener().  I
> think this would pop out all of the listening sockets quickly from the
> listener pollset.

How would you signal it?

> Solution: Actually start working on 3.0 with a sane event/start/shutdown
> model?

That'd be nice alright :-)

Colm MacCárthaigh                        Public Key:

View raw message