I have 2 nodes with one agent on each of them. I have 1 process that runs on node 1, but if node 1 is restarted it will run on node 2. Dynatrace detects this process like 2 processes in 1 process group and if I want to see the process availability % in Data explorer it shows 50% availability. I wait about 3 hours and this didn’t change.
Is there some timeout after which the process will not be monitored anymore? Is there some configuration about this? I saw that the process disappears from Hosts view after about 5 minutes inactivity.
Hi,
No, for now I’m trying to understand how Data Explorer works. I expected that if I want to check the process availability % in last 10 minutes, since there is only one process in this process group in the last 10 minutes to see 100% availability. Instead of that I see 50% availability, because the Data Explorer still “remember” that a long time ago (3-4-5 hours) there was a process with the same name on some other host. So I want to understand where is the mistake in my logic?
I see the same behavior on the demo environment. Since the metric used is:builtin:pgi.availability, my best guess is that by default the 72 hours is used that also applies for the smartscape topology. But perhaps we can get someone from dynatrace to confirm this.