How can we setup a Incident to alert on Heap usage of JVM still being above 85% after some time after a full GC has occurred . ( in short a High JVM heap after a full GC)
I don't think this is possible exactly as you have requested but you can design an incident that does basically the same thing. You can create a memory utilization measure with an upper threshold of 85%, then you can add it to your incident with the aggregation set to average. At this point, you would choose an evaluation timeframe that is large enough to allow for a full GC but not too large that your memory would be in a high utilization without state without you knowing, say five minutes. This should alert you if the average memory utilization was 85% or above for the past 5 minutes. You can tweak to allow for more time if needed but hopefully, that provides you with the information you need.