[NTLUG:Discuss] apache error message

[David Camm]

Victor Brilon wrote:
> 
> You'll see messages like this if  Apache has too many process already
> running (i.e. a bunch of CGI's running that haven't exited) and they eat up
> all the available file handles that the OS can handle, 

i currently have file-max set to 4096. would making this larger (like
6144) possibly help?

>or the process table
> gets full for the parent process (i.e. Apache). Check to see that some long
> running processes aren't doing this. You'll also get this message if a
> process has eaten up all the memory and swap space and there's no enough
> left for a new process to be forked off and executed.

although i have 1Gig RAM on the machine, being out of memory IS a
possibility....

> 
> Victor
> 
> At 02:19 PM 3/5/2001 -0600, David Camm wrote:
> >this is the second time i've noticed error messages linke this in my
> >logs (apache 1.3.9):
> >
> >[Mon Mar  5 13:13:40 2001] [error] [client 12.75.65.208] (11)Resource
> >temporarily unavailable: couldn't spawn child process:
> >/usr/local/etc/httpd/cgi-bin/byrd/rotate7.pl
> >
> >they appear for about 1 to 1-1/2 minutes, always referencing cgi
> >programs, and then go away.
> >
> >could i have exceeded the maxclient number in httpd.conf (i tend to
> >doubt this)? or is this a kernel issue?
> >
> >can anyone help me by explaining why this is happening, and what if
> >anything i can do to stop it from happening again?
> >
> >as an aside, does anyone know where theses messages are documented? they
> >are NOT in the o'reilly apache reference, nor have i been able to find
> >them on the web.....
> >
> >thanks!!!
> >
> >david camm
> >advanced web systems
> >keller, tx
> >_______________________________________________
> >http://ntlug.org/mailman/listinfo/discuss
> 
> _______________________________________________
> http://ntlug.org/mailman/listinfo/discuss