We pretty new to dynatrace and have been testing doing some host monitoring using host agents.
It seems from my test that agent groups don't actually work as the target of a measure for a host monitoring rule.
here is what I did:
make a new disk space % measure (>1% warning >5% criticial). This is just to generate some incidents and data we can graph. Leave target as default (all agent groups)
It goes off like you would expect and you can see the disk % on a chart
Edit the measure so it only targets some agent groups say our one we created for windows hosts agents only and save that & incidient rule.
Result - all graph data goes blank so it no longer seems to be collecting anything. If I delete the incidents they don't come back for just the agent group. I tried a few agent groups all the same result (even ticking all the agent groups that exist)
go back into measure and tick it back to default 'all agent group' and save.
hey presto graph data reappears and incidents come back.
Answer by Andreas G. ·
Can you let me know which measure you are exactly talking about. Is it the Host Performance -> Disk Storage Utlization? Or is it a different measure? The Disk Storage Utilization measure is one delivered through our host monitoring capabilities of our agents. This measure can only be configuired to either collect data for all agents or none - it doesnt provide a "per agent group" option. Thats why I am wondering what type of measure you are trying to use so that I can verify this behavior.
If what you say is correct I assume this is a bug
Answer by Anton G. ·
This is a bug and a huge headache for us as well.. Although you can specify the agent groups it only stops the measure from collecting data and does not actually break the data into the selected Agent Groups. There is a RFE for this Enhancement - Agent Group Calculation to Host Performance Metrics
Support number 00873948