diaglogd is a resource hog.

[Gary Glick]

Recently on one of my servers UX servers the diaglogd daemon has started utilizing about 95% of the cpu according to the "top" display. It is interfering with the database operations on the server. The catch is I can't figure out what it's upset about. It doesn't appear to be logging any errors.

Disks & memory are clean. The last entry reported by logtool is from 2003. Nothing in syslog.

I'm not sure what to do next, I've had to turn of diaglogd so the DB users can do their work.

Thank you for any suggestions

[Pete Randall · Accepted solution]

Gary,

Here are a couple of items from the TKB that may help:

http://www1.itrc.hp.com/service/cki/docDisplay.do?docLocale=en_US&docId=200000079998988
http://www1.itrc.hp.com/service/cki/docDisplay.do?docLocale=en_US&docId=200000079996717


Pete

[Gary Glick]

Got it fixed.
stopping diagnotics
Moving the /var/stm/data/diaglogd_hold_list
file out of the way then restarting the diagnostics fix the resurce hogging problem.

Thank you

Gary

[Andrew Merritt_2]

What version of OnlineDiags have you got installed? That problem applied to a very old version; if you're still seeing that with a recent version I'd be interested to know more.

If you're using an old version, you need to upgrade to something recent.

Andrew