Status Unknown

I admit it; it can sometimes be bewildering and frustrating understanding the ‘unknown’ status APM gives. You’ve posted as much on several thwacks and, well, I’ve got some explaining to do. First, a brief review on how another Orion product works.

NPM for example, uses ICMP to determine the status of a node by sending pings in a process that works something like this:

“Ping”, echo response = node is up.

“Ping”, no response = node is down.

APM doesn’t ping to determine the status; it polls using the scheduler. On Windows servers, it looks at processes that are running by querying the Windows Management Instrumentation (WMI). As a point of clarity, there are SNMP monitors for Windows servers too; but here we are focusing on WMI monitors. If for some reason WMI is incorrectly configured or doesn’t respond in a timely fashion, performing another query doesn’t always help. In fact, WMI itself could have a problem while the application is fine. The status is reported as Unknown until APM is certain it can report the true health. The component status text (shown in the little orange info box on the App Details view) will tell you why the component status is unknown.

I hope this explanation helps you to understand how APM determines the status. An alternative solution may be in gaining access to the state of processes without going through WMI. We are actively investigating all options to address this in a timely manner.