On Saturday 07 February 2004 06:31 pm, Erik Bourget wrote: > Hello; > > I've got clamd processing a ton of mail, it does a good job not crashing > these days (cvs as of a week or so ago), but the new problem is as bad or > worse - the hanging. At least when it crashed, supervise (I'm running it > under daemontools) could restart it. Does anybody have a clean way to > detect clamd failure and kill it if it happens? Ideally this would be a > process that I could put in between supervise and clamd that would exit on > clamd failure. > > Barring a canned solution: > I've got clamd on a system. How do I detect that it is hung, apart from > "clamdscan never returns"? Magic with the .ctl file?
I'm using a program called "monit" (http://www.tildeslash.com/monit/), and it works great. It connects to the port and makes sure it's open, checks if the process is a zombie, and does some memory and load avg. checks. Here's my monit config snippet for clamav: check process clamd with pidfile /var/run/clamd.pid start program = "/etc/init.d/clamav start" stop program = "/usr/bin/killall -9 clamd" if failed host 127.0.0.1 port 3310 then restart if memory is greater than 5% then restart if loadavg(1min) is greater than 10.0 then restart alert [EMAIL PROTECTED] If you're using Unix sockets, it'll connect to that too. -- Matt Systems Administrator Local Access Communications 360.330.5535 ------------------------------------------------------- The SF.Net email is sponsored by EclipseCon 2004 Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA. http://www.eclipsecon.org/osdn _______________________________________________ Clamav-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/clamav-users
