A technical post here… this weekend we had a problem where one of the partitions on our zimbra email server reached 100% capacity, and resulted in us doing a hard restart.  As a consequence of this on restart we could not get zimbra MTA to start, the problem being that postfix thought it was already running.  The solution for anyone who finds this is  as follows, pasted to gether from some notes from “Dave”, who fixed it. “The problem was that whenever postfix is started, it runs postalias. Postalias takes files like /etc/aliases and translates them into machine-readable files.  There was an aliases.db file hanging round in /etc, owned by root. When zimbra tried to start up wuth the zimbra id, it couldn’t touch the stale file.  Unfortunately, this hung it rather than erroring out cleanly. I cleaned up the stale files, and started zimbra using root and /etc/init.d/zimbra start. That reset the necesary permissions and recreated the files.”.  I hope this helps if you are a Zimbra user that finds that same problem.