I have worked around the database performance problems by deleting and
recreating the events database and then setting the "Delete Historical Events
Older Than (days)" setting to 90 days. I had over 5,000,000 events in the
history table before doing this. I hope this will prevent the database from
getting so big that it again becomes a problem.
So far, the Zenoss GUI is much snappier (very small events database) and the
server perf stats are healthier (much less average CPU). There is also no
ZenDeleteHistory.py error or CPU nasties daily at 11 AM.
Here are the steps I took (this completely clears out the events history):
1. Drop the events database (logged in as root):
mysqladmin -prootpassword drop events
2. Create a new events database (logged in as user zenoss)
zeneventbuild localhost zenoss zenossuserpassword events
3. Use GUI SQL Administrator installed on the Zenoss server to create
additional database user permissions as needed to run remote SQL Administrator
(for convenience) on my Windows XP machine.
I also am trying to modify the "/Events/Archive/defaultmapping" event class to
try and keep some nuisance debug events (like /Perf/SNMP "Error reading
value...") from getting archived over to the history database table. Not sure
yet if this is giving me the results I want.
Matt
CentOS 5.2
Stack-installed Zenoss 2.4.2
-------------------- m2f --------------------
Read this topic online here:
http://forums.zenoss.com/viewtopic.php?p=38800#38800
-------------------- m2f --------------------
_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users
