I have been working through our environment to try to cut down on the chatter for our windows server alerts. Just to take two items that always come back i have a basic template assigned to my windows servers that monitors Total Available MB and % Processor Time (performance counters) among other things.
I currently have the template set to use baseline thresholds and they are set to go to a warning for 5 of 8 polls and then critical for 3 of 5 polls (i currently poll the servers every 5 minutes). Even after going back and manually recalculating baselines and in some cases looking at the historical graphs and manually setting thresholds i still receive a lot of junk alerts.
So the question that's been asked a thousand times I'm sure, and has a different answer each time. What is everyone else doing to set their thresholds? Balancing polling types / # of polls and the threshold values to receive good alerts?