SNMP & Logger don't start in scripted restart

[dwmtractor]

Hey everybody, this is slightly like some things others have mentioned, but different enough that I figured I'd throw it out as a separate thread:

When I upgraded from 5.08 to 5.011 on Ubuntu 6.06 LTS, everything started normally except for (1) a failure of Clam due to newer version (fixed), and (2) SNMP and Logger failed to start at the script end. Once I fixed clamAV and restarted Zimbra everything came up so I assumed (wrongly, it now seems) that everything is OK.

I use the shell script backup method (the really simple version, pre some of the enhancements others have made), and part of the script is an su stopping and starting of Zimbra services. When they restart from the script, SNMP and Logger are stopped again. However if I log onto the command prompt and manually zmcontrol stop and start, they come up flawlessly.

What is different about calling these same commands from a shell script, that causes this failure? Where in the logs do I even begin to look for errors?

Thanks in advance for any suggestions!

[dwmtractor]

Follow-up: I was using my own shell script for backup, and in prior versions I had found that the command

sudo -u zimbra /opt/zimbra/bin/zmcontrol start

was the only one that worked for me on Ubuntu with Zimbra 4.5.x. However, it seems with the newer version something about the way sudo calls the programs blows them up. I created a test batch for the sole purpose of stopping and starting the services, and watched them fail.

The alternate syntax:

su - zimbra -c "/opt/zimbra/bin/zmcontrol start"

did the job.

I will watch the cron'ed run tonight before declaring victory, however. . .