What should your total time/threshold be? Let AppNeta Performance Manager (APM) run for a bit to establish a baseline and then set your thresholds around that.


Once you get APM going, if you find that you’re consistently operating outside the satisfied zone, your alert profile is notifying you that performance is out of tolerance, as opposed to performance being poor (it’s already poor). This info is still valuable for responding to production incidents, but you should be focusing on optimizing.


Best case scenario, APM proves that your app is already in good shape, in which case your alert profile is notifying you about when performance is poor. Changes in performance could be because of recent code changes, a bug, or insufficient capacity during events or peak season.


Regardless of which Apdex zone you’re operating in, what you want to do is establish a baseline for performance and then set your threshold just above that. It’s expected that you might need to adjust this. Too close to your baseline, and you’ll be alerted too frequently about transient conditions, too far way and there will be a delay in your incident response time.