On 10/22/2014 2:45 PM, steve wrote: > > > Hi, > > This weekend a MySQL server which backends a couple of postfix servers went > down for a couple of hours without anyone noticing and I'm looking to setup > some monitoring so that it doesn't go unnoticed again in future. >
There are really only two good options for monitoring postfix. + base level monitoring -- is postfix master process running? bonus points: see if postfix answers with a 220 response on the smtp port. + operational monitoring -- a process submits a message periodically, another process monitors the target mailbox to see if a message arrives periodically. Anything in between is pretty meaningless. In the case of a broken lookup table interrupting receiving mail, only "operational monitoring" would reveal that.