www-apache-bugdb mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Dean Gaudet <dgau...@arctic.org>
Subject Re: os-linux/457: apache root server process goes crazy and sucks down all available cycles
Date Wed, 23 Apr 1997 00:00:02 GMT
The following reply was made to PR os-linux/457; it has been noted by GNATS.

From: Dean Gaudet <dgaudet@arctic.org>
To: Brent Phillips <brent@jamisongold.com>
Subject: Re: os-linux/457: apache root server process goes crazy and sucks down all available
cycles
Date: Tue, 22 Apr 1997 16:55:09 -0700 (PDT)

 
 Are you invoking httpd multiple times and using Listen?  If so have you
 moved your ScoreBoardFile somewhere different for each server? 
 
 You can get around that last thing (which we never documented until 1.2
 docs) by adding -DHAVE_SHMGET to your EXTRA_CFLAGS in your Configuration
 file and rebuilding.  Give that a try. 
 
 Otherwise the only thing we can ask is to try 1.2b8 (or b9 in a short
 while) as we've fixed many related bugs. 
 
 Dean
 
 On Tue, 22 Apr 1997, Brent Phillips wrote:
 
 > 
 > >Number:         457
 > >Category:       os-linux
 > >Synopsis:       apache root server process goes crazy and sucks down all available
cycles
 > >Confidential:   no
 > >Severity:       serious
 > >Priority:       medium
 > >Responsible:    apache (Apache HTTP Project)
 > >State:          open
 > >Class:          sw-bug
 > >Submitter-Id:   apache
 > >Arrival-Date:   Tue Apr 22 16:40:01 1997
 > >Originator:     brent@jamisongold.com
 > >Organization:
 > apache
 > >Release:        1.1.3
 > >Environment:
 > Linux 2.0.30, gcc
 > >Description:
 > We're running about 10 different virtual servers on an Intel Linux 2.0.30 
 > machine. Occassionally, one of the root apache processes goes crazy and 
 > sucks down all available processor cycles; it has to be killed for the system
 > to continue normal operation. We haven't seen this happen with the child 
 > processes - it always seem to be the root process. 
 > 
 > This problem started occuring when we added a virtual server with a moderate
 > to heavy load, so it may be related to using up some limited system resource.
 > >How-To-Repeat:
 > Run several web servers on an Intel Linux 2.0.30 and have robots hit them
 > mercilessly; wait until the system gets unresponsive and check on the web
 > server processes' CPU usage.
 > >Fix:
 > I would guess that some resouce is running out
 > >Audit-Trail:
 > >Unformatted:
 > 
 > 
 > 
 

Mime
View raw message