SO_LINGER (was Re: Style list summary)

[Chuck Murcko <ch...@telebase.com>]

Rob Hartill liltingly intones:
> 
> > 
> >   Our HPUX 1.1b5-dev  (xxx.lanl.gov) has stopped accepting connections
> >   twice recently.. needing a SIGHUP to bring it back. 
> > 
> >   Any ideas on what would cause that?
> > 
> > Hmmm... did you ever see this before the SO_LINGER patch?
> 
> We've had things freeze in the past but they are very rare and could
> be caused by any number of things outside of Apache's control, but
> since switching to b5-dev the server has died twice (Saturday and today).
> 
> The machine is low use (~40,000 hits/day) with lots of timeouts from
> distant clients. We SIGHUP once a week.
> 
> We now have a cron script to look for an outage and SIGHUP automatically.
> 
> My guess is that if the SO_LINGER patch is newer than a couple of weeks
> then it could be a factor. If it's much older than that then I doubt it.
> 
It went in on 27 May, for 1.1b4, I believe.

chuck
Chuck Murcko	N2K Inc.	Wayne PA	chuck@telebase.com
And now, on a lighter note:
There are three things I always forget.  Names, faces -- the third I
can't remember.
		-- Italo Svevo