I know I'm running an older version of Zimbra on a now unsupported platform, but I just wanted to provide some feedback on a curious issue I had noticed. Zimbra is actually working normally, all messages are being delivered properly, but the following error appears with alarming regularity in /var/log/mail.log and /var/log/zimbra.log:
postfix/postqueue: fatal: Queue report unavailable - mail system is down
After some searching and reading older discussions here, I realized that OS X includes Postfix as its MTA, and the binaries are located in /usr/bin and /usr/sbin.
mail:/var/log user$ which mailq
mail:/var/log user$ which postfix
mail:/var/log user$ mailq
postqueue: fatal: Queue report unavailable - mail system is down
Switching to the Zimbra user produces the correct results:
mail:/var/log user$ sudo su - zimbra
mail:~ zimbra$ which mailq
mail:~ zimbra$ which postfix
mail:~ zimbra$ mailq
-Queue ID- --Size-- ----Arrival Time---- -Sender/Recipient-------
4EA1A3D5A5C 5383 Thu Mar 15 07:19:12 MAILER-DAEMON
(connect to ********: Operation timed out)
-- 5 Kbytes in 1 Request.
The only conclusion I can draw, then, is that for some reason the process that calls Postfix's mailq for the system log report doesn't have the right $PATH in its environment, and as a result it's calling the system mailq binary.
Any suggestions on where I can locate and/or fix this so Zimbra logs properly?