Quote:
Originally Posted by scottf76 Just want to be thorough. So besides the metrics / status info in my first email is there anything else that you can think of?
thanks. |
I'm not sure how to respond...
When we set up monitoring for any server, we approach the question of "what to monitor" by listing what services, daemons, etc. which need to be functional for the end-user(s) to deem that the server is "working OK".
In creating that list, we document use cases of how the user(s) will be interacting with the system, and whatever they will rely on, we will monitor. If what they rely on has prerequisite dependencies, we'll monitor those, too.
Think about things like the number of emails in the Deferred queue for Postfix. Some number is normal. But if the number keeps increasing you might have an amavis problem, and since amavisd is frequently starting and destroying processes, there's not one or a few non-changing amavis processes you can monitor.
Too much monitoring isn't as bad as too little, but in our view, "less is more".
Does that help?
All the best,
Mark
__________________
___________________________________
L. Mark Stone, CIO
"Uptime. All the time."
477 Congress Street | Portland, ME 04101-3431 | (207) 772-5678
proactive maintenance and monitoring | technology consulting
Zimbra groupware | EMR implementations | private cloud hosting