stopping service status messages

[timwiel]

How does one stop or at least slow down the service status messages being sent to the admin account?

Subject: Service stats started on gideon.XXXXXX.XX.nz

Body:

Feb 26 15:32:15 gideon zimbramon[565]: 565:err: Service status change: gideon.XXXXXX.XX.nz stats changed from stopped to running

[timwiel]

bump bump bump

[madscientist159]

Did anyone ever find a solution to this problem? My admin mailbox was flooded with over 25000 messages in just a couple of months, while Zimbra works perfectly in all other respects. They seem to come in at a rate of about 100-150 per hour.

Thanks!

EDIT: I found a temporary workaround that might be helpful:
1. Edit (as root) /opt/zimbra/conf/localconfig.xml
2. Change the smtp_notify key value to read no instead of yes
3. Restart Zimbra (as the zimbra user)

Of course it would be best to understand why this flood is occurring, but this will at least stem the tide and hopefully reduce loading on the Zimbra mail services. My hunch is that my server is just too slow to respond with a complete service status in a timely manner, as the number of Emails had increased after a secondary CPU failure and subsequent reduction in computing power.

[madscientist159]

I made some headway in actually shutting off the spigot at the source.

Apparently, the swatchd process uses a nearly continuous tail of /var/log/zimbra.log to generate those messages. Problem is, if the /var/log/zimbra.log file is not rotated out properly (or deleted) swatchd just keeps sending and sending, ad nauseum.

Now, I don't know how to fix the log rotation problem, but clearing the /var/log/zimbra.log file in addition to the hack mentioned above seemed to work around the problem on my server. Whether this is acceptable in your environment is up to you, but consider that my server was bogged down with 450,000 messages an hour before the workaround was applied.

Any thoughts on the log rotation problem from the Zimbra devs?