httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Marc Slemko <ma...@znep.com>
Subject Re: server timeouts and FIN_WAIT_2
Date Wed, 08 Jan 1997 05:37:32 GMT
Hmm.  This does seem to be something which more than one person has
observed happening in 1.2 but not 1.1.1.  Or are people seeing a bigger
difference between 1.0 and 1.1?  That could be a lot of things, possibly
keepalives.  (and from a quick look just disabling keepalives won't make
the code do the same thing it would in the pre-keepalive code).

Try disabling the call to sock_disable_nangle and any SendBufferSize you
have.  

The only other change in http_main.c that stands out is the hard_timeout
and the reap_children changes.

OTOH, Chuck may be seeing a different thing than others.  Unfortunately, I
can't yet use 1.2 on our server that gets a good number of hits to see if
I could see the problem here. 

On Tue, 7 Jan 1997, Chuck Murcko wrote:

> No difference without lingering close. OTOH, my internal proxy's FIN_WAIT_2
> state connections are all from WIN95 boxes. 8^)
> 
> Chuck Murcko liltingly intones:
> > 
> > Oh, right. Crap. I can try this. Will do now, in fact.
> > 
> > Marc Slemko liltingly intones:
> > > 
> > > lingering close; could be an obscure implementation bug in apache code or,
> > > more likely, bugs in various OSes that don't do a half-close properly. 
> > > Has anyone with the problem tried compiling with NO_LINGCLOSE to see if
> > > it goes away?
> > > 
> > > On Tue, 7 Jan 1997, Chuck Murcko wrote:
> > > 
> > > > Is there something specific about the way Apache handles timeouts and/or
> > > > connection closes now that could be causing connections left in FIN_WAIT_2?
> > > > I can't see anything obvious in the source, but these are areas of the
> > > > code I'm not intimately familiar with.
> > > > 
> > > > Did the basic way timeouts or closes are handled by the server change
> > > > between Apache 1.1.1 and Apache 1.2 change in any way?
> > > > 
> > > > We seem to have some folks in c.w.i.s.u. who are convinced that this is
all
> > > > a problem with Apache 1.2.
> > > > 
> > > 
> > > 
> > 
> 
> chuck
> Chuck Murcko	N2K Inc.	Wayne PA	chuck@telebase.com
> And now, on a lighter note:
> Every improvement in communication makes the bore more terrible.
> 		-- Frank Moore Colby
> 


Mime
View raw message