Google for "deadlock debugger" and "debug spinning zope". -PW On Wed, Jan 24, 2007 at 03:11:52PM -0500, Timothy Ball wrote:
I have a zope server (2.7.4) that I maintain and every now and then the python process that's running zope goes heywire and eats all of the cpu resources and basically make my website stop responding. Does anyone have any clues as to how I can debug this problem?
I'm not the best python coder but I do know how to use pdb. And I'm looking for sorta where and what pieces would be the best place to start putting the trace() bits.
... but really any sort of guidance would help... here's a small bit from ps auxwwww showing a bit of what I mean, but when it *really* goes heywire it just eats all of my cpu and causes the load to shoot up near the 50s (on a dual opteron 275 processor machine)
USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND zecms 4062 12.6 8.9 830640 718768 ? Sl Jan19 972:32 /thingie/src/bin/python /thingie/src/opt/Zope2/lib/python/Zope/Startup/run.py -C /www/zed/src/var/zope/etc/zope.conf
TIA, --timball
_______________________________________________ Zope maillist - Zope@zope.org http://mail.zope.org/mailman/listinfo/zope ** No cross posts or HTML encoding! ** (Related lists - http://mail.zope.org/mailman/listinfo/zope-announce http://mail.zope.org/mailman/listinfo/zope-dev )
-- Paul Winkler http://www.slinkp.com