Received: (qmail 36547 invoked by uid 501); 25 Jun 2001 17:46:39 -0000
Message-Id: <20010625174639.36546.qmail@apache.org>
Date: 25 Jun 2001 17:46:39 -0000
From: Valerio Morettini <valerio@wnet.it>
Reply-To: valerio@wnet.it
To: submit@bugz.apache.org
Subject: Apache eats 100% CPU and memory, getting unusable.
X-Send-Pr-Version: 3.110

>Number:         7929
>Category:       os-linux
>Synopsis:       Apache eats 100% CPU and memory, getting unusable.
>Confidential:   no
>Severity:       critical
>Priority:       medium
>Responsible:    apache
>State:          open
>Quarter:
>Keywords:
>Date-Required:
>Class:          sw-bug
>Submitter-Id:   apache
>Arrival-Date:   Mon Jun 25 10:50:02 PDT 2001
>Closed-Date:
>Last-Modified:
>Originator:     valerio@wnet.it
>Release:        1.3.20
>Organization:
apache
>Environment:
Linux 2.4.6pre3 on a double PIII 600 Mhz, 1 Gig RAM
glibc-2.1.3
egcs-2.91.66
apache 1.3.20 + php4.0.6 + mod_rewrite

Compiled-in modules:
  http_core.c
  mod_log_config.c
  mod_mime.c
  mod_include.c
  mod_dir.c
  mod_cgi.c
  mod_alias.c
  mod_rewrite.c
  mod_access.c
  mod_auth.c
  mod_setenvif.c
  mod_php4.c

Server version: Apache/1.3.20 (Unix)
Server built:   Jun 22 2001 19:03:30
Server's Module Magic Number: 19990320:10
Server compiled with....
 -D HAVE_MMAP
 -D HAVE_SHMGET
 -D USE_SHMGET_SCOREBOARD
 -D USE_MMAP_FILES
 -D USE_SYSVSEM_SERIALIZED_ACCEPT
 -D SINGLE_LISTEN_UNSERIALIZED_ACCEPT
 -D HTTPD_ROOT="/usr/local/apache-normal"
 -D SUEXEC_BIN="/usr/local/apache-normal/bin/suexec"
 -D DEFAULT_PIDLOG="logs/httpd.pid"
 -D DEFAULT_SCOREBOARD="logs/httpd.scoreboard"
 -D DEFAULT_LOCKFILE="logs/httpd.lock"
 -D DEFAULT_XFERLOG="logs/access_log"
 -D DEFAULT_ERRORLOG="logs/error_log"
 -D TYPES_CONFIG_FILE="conf/mime.types"
 -D SERVER_CONFIG_FILE="conf/httpd.conf"
 -D ACCESS_CONFIG_FILE="conf/access.conf"
 -D RESOURCE_CONFIG_FILE="conf/srm.conf"
>Description:
After some hours that the server is running, it starts eating up all available cpu, quickly making the web server unusable. When i try to stop it, i get ALOT of messages of the kind "Child process xxx still did not exit..sending a SIGTERM" and below ALOT of lines like "Child process xxx still did not exit..sending a SIGKILL". I really don't know how to send a backtrace, since the system reaches such a cpu load that anything stops working, even telnet and modem connections. One thing i've noticed that maybe can help...i see that normally there are about 500 processes in the system, with 1 to 5 in "Running" state...one thing i noticed during the problem, is that there are MANY processes in "running" state, even 40 or 50, with the same number of total processes (500 or so)...might this be a kernel issue? The only thing to do to prevent crash is stopping and restarting apache every few hours...which is not good at all i guess...
>How-To-Repeat:
It seems to happen randomly, but heavy load causes it to happen more often.
>Fix:
i wish i had :P
>Release-Note:
>Audit-Trail:
>Unformatted:
 [In order for any reply to be added to the PR database, you need]
 [to include <apbugs@Apache.Org> in the Cc line and make sure the]
 [subject line starts with the report component and number, with ]
 [or without any 'Re:' prefixes (such as "general/1098:" or      ]
 ["Re: general/1098:").  If the subject doesn't match this       ]
 [pattern, your message will be misfiled and ignored.  The       ]
 ["apbugs" address is not added to the Cc line of messages from  ]
 [the database automatically because of the potential for mail   ]
 [loops.  If you do not include this Cc, your reply may be ig-   ]
 [nored unless you are responding to an explicit request from a  ]
 [developer.  Reply only with text; DO NOT SEND ATTACHMENTS!     ]