We configure maintenance windows for our environment, to make sure we dont report on data captured when our environment is impacted by planned work. At least that is what I thought...
We have a method response time (builtin:service.keyRequest.response.time) that management specifically require that we report the average value of. We do this on a dashboard with a single value tile, that illustrate the average of the method response time.
Last weekend we had a maintenance window configured, and when I looked at our single value dashboard tile on monday, the average response time for the period was way higher than normal. Turns out, that for whatever reason, the method was captured 3 times during the maintenance window with a response time of 59 minuttes.
It turns out that Dynatrace does not filter out data captured during maintenance windows for this metric, as it does with user actions & synthetic monitors.
This really baffles me, and I discussed the issue with the chat - who said that there was no option to exclude recorded data, making it rather hard to report a correct number for management.
How does other Dynatrace users deal with such issue?
Dynatrace is always looking at the data, the MW has 3 options with the most in-depth option just suppressing the alerts and disabling the detection of the problems. The metric value will still be there but it wont be listed as a "Problem" Some tiles offer the ability to exclude data from the maintenance windows. Sounds like a great RFE.