04 Jan 2021 04:53 PM
Hi!
We discovered today that our failure rate anomaly detection based on a custom metric was frequently higher than 100%. Can someone explain me why?
Thank you...
Solved! Go to Solution.
04 Jan 2021 07:14 PM
It appears that you have the Unit set to Percent, and the aggregations are set to auto, sum. I suspect that the system is looking at your percentages then adding them up to give you a value over 100%
05 Jan 2021 09:55 AM
I've tested with Failure rate as seen by caller, and it seems much more adequate:
The sampling is not the best, there are few accesses, but values seem far better now.
Thank you!