httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Brian Behlendorf <>
Subject Re: [Fwd: Re: socket error on APACHE v1.0]
Date Sun, 11 Feb 1996 01:53:14 GMT

Are we talking about "socket error: accept failed" in the error logs?  Yep,
I'm getting them too on Solaris 2.5, in the main error log (whereas "send
timed out for" will also show up in the virtual host error logs as well). 
Looking at the error log it appears to happen even at really unloaded 
times, like 5am, so I dunno that slow OS response can be blamed.  
I'm still running 1.0, I'll be installing 1.0.2 today and we'll see if it
still happens.   I don't see it on hyperreal.


On Fri, 9 Feb 1996, Aram Mirzadeh wrote:
> > > > This could very likely be caused by the sending of SIGHUP once per day.
> > > 
> > > No, it doesn't.  Apache just has this problem.  And we're not addressing it
> > > why I don't know. 
> > 
> > We can't address what we can't see. Some of us simply don't get these
> > problems. They seem to be platform dependent, with linux being the
> > biggest sufferer. An earlier user explanation I forwarded seemed to
> > blame slow OS responses to a lost connection that had just been
> > negotiated and was awaiting a process to accept it.
> I've been telling everyone on this list, about this problem since .8.6....
> It has been getting worse, and worse as our version go up, and the processes
> are getting more complicated.  The problem is there, maybe it's linux only
> maybe it's apache only, but apache seems to be the ONLY httpd server that has
> the problem, the others don't.  That means, there is either a way around it
> or we're doing something wrong. 
> > 
> > If you think more can be done, what do you suggest?
> I don't know, what do you need to fix it?  You tell me what you, or any 
> other programmer needs, and I'll see about getting it to you... 
> <Aram>
> -- 
> #define BITCOUNT(x)     (((BX_(x)+(BX_(x)>>4)) & 0x0F0F0F0F) % 255)
> #define  BX_(x)         ((x) - (((x)>>1)&0x77777777)                    \
>                              - (((x)>>2)&0x33333333)                    \
>                              - (((x)>>3)&0x11111111))
>                 -- really weird C code to count the number of bits in a word
> Aram Mirzadeh

--=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--  http://www.[hyperreal,organic].com/

View raw message