Date: Tue, 19 Sep 2006 00:58:16 -0400 From: "Darrell D. Mobley" <dmobley (at mark) uhostme.net> Subject: [coba-e:07060] Odd events for September 19, 2006 To: <coba-e (at mark) bluequartz.org> Message-Id: <000301c6dba8$387a0fb0$6400a8c0@YOUR4105E587B6> X-Mail-Count: 07060Well, this getting to be a daily thing for me. Ok, tonight I am using my website which runs on my dedicated BQ server and notice it grind to a halt. The website is a Joomla CMS system which runs in PHP and MySQL. There were approximately 3 users on the site. The server is a P3-500Mhz Dual Processor with 512MB RAM. So I SSH into the server and watch the load climb up as high as 119 before I finally am able to stop the mysqld and httpd services. This took approximately 20 minutes to accomplish as well as allow it time to return to normal load before I restarted the services. In /var/log/httpd/error_log, there are 120 of these entries with differing child process numbers: [Tue Sep 19 00:23:35 2006] [warn] child process 8009 still did not exit, sending a SIGTERM When things started going downhill, obviously MaxClients gets exhausted and the system begins to swap memory, further exacerbating the problem. My question: is this situation normal or should I suspect something else. I did notice two cron jobs running, one of which was break-smrsh-loop.pl and wondered if perhaps sendmail was looping: crond \_ crond | \_ /bin/sh -c /usr/local/sbin/break-smrsh-loop.pl > /dev/null 2> | \_ /usr/bin/perl /usr/local/sbin/break-smrsh-loop.pl | \_ /usr/bin/perl /usr/sausalito/sbin/get_quotas.pl \_ crond \_ /usr/bin/perl -I/usr/sausalito/perl /usr/sbin/swatch -c /etc/ \_ /bin/sh /usr/sausalito/swatch/bin/am_dns.sh \_ /usr/bin/host -W 2 127.0.0.1 127.0.0.1 \_ grep 1.0.0.127.in-addr.arpa. Maybe I should just put up a blog where you all could subscribe to my daily trauma. At least my access_log hasn't zeroed itself out yet, but that shouldn't happen for another 4 hours.7060_2.html (attatchment)(tag is disabled)