Difference between connectivity and availability metric for a process in Dynatrace

[suresh230591]

Hi,

Can someone please elaborate the difference between process availability and process connectivity in Dynatrace. Can we consider process connectivity metrics an equivalent to availability.

I could not find a metrics related to process availability. However, Dynatrace alerts on when a process goes down through anomaly detection.

Thanks

Suresh

[skrystosik · Accepted solution]

Process availability is measured by oneagent. If process is working it's available. Connectivity is network metric. If other processes can connect to investigated process, connectivity measure is 100%. It's also related to TCP connections, more detailes here:

https://www.dynatrace.com/support/help/shortlink/n...

Regards,
Sebastian

[wolfgang_beer]

Great answer, thx. You were 5 seconds quicker 🙂

[wolfgang_beer · Accepted solution]

Process availability is collected by our OneAgent on OS level, which means if the process is up and running. It does not tell you anything about the process being responsive or correctly doing its job.

You can fetch the process availability by using the timeseries id shown within the screenshot below:

[suresh230591]

Is this timeseries Id also available for custom charting. Is there way we can show this on dashboard.?

[skrystosik]

I don't think there is any actually. Such things are reported as events (restarts, deploy changes etc). It's different channel of metrics.

[suresh230591]

Thanks Sebastian and Wolfgang for your quick help

It would be great if we can have this metric for charting as well. We have this requirements from the customer, they want to see the availability of critical application processes on dashboard. Like we have traffic light tile for hosts and services where we immediately come to know about any problem related to services and hosts of an app just by looking at the dashboard.

Thanks

Suresh

[skrystosik]

Generally process inavailability without maintenence window will also be rised as problem

[EXTERN_PUPPE_MI]

I would like to see this metric for charts as well. I would expect timeseries used in the process/host view to be available in custom charts in general.