Seems to me those kind of issues would be a good way to test your monitoring. For instance verifying that the appropriate people are notified or automatic actio

Seems to me those kind of issues would be a good way to test your monitoring. For instance verifying that the appropriate people are notified or automatic action is taken when within a reasonable amount of time after the CPU on the box starts spiking.