Windows Service Availability

[djmnyc]

Good afternoon.

I have what I believe to be a simple ask of an observability/monitoring tool, but I am finding it very difficult to setup within Dynatrace.

Very simply, I need to ensure on a set of production application servers that two separate tomcat service instances are always running - tomcat and tomcat-RS.  I was told out of the box Dynatrace automatically monitors certain key processes, including tomcat.  However, I found out a problem is only reported on a host running tomcat if the service is forcefully shutdown.  If someone manually shuts the service down, this is a graceful shutdown, and no problem is reported.  This is not what we need.

From my initial research and asking around, I found two possible solutions:

1. Settings/Processes and Containers/Process Availability

2. Settings/Monitoring/OS Services Monitoring

For my subset of production application server hosts running tomcat, there is no assigned host group, but we have tagged them in a meaningful way to easily filter/select the subset.  For example, we have created the following tags to identify our production application servers for this one product that uses tomcat:

ENV_PROD:1, APP_DMS:1, FUNC_APP:1

How would someone go about setting up a service availability rule in Dynatrace so that whenever either one of the two tomcat instances is not running for more than X minutes (on hosts with the tags above) a problem is raised (which then initiates an alert email)?  

Seems pretty straightforward, but I cannot get a simple answer.  Chat support just sent me to the OS services monitoring page, which as I stated above does not give examples on how to limit which hosts you are checking on.

Can someone please explain how they would set this up given the above scenario?

1. Is OS service monitoring the best way to do this, or would you use Process Availability?

2. Is it better to assign host groups instead of using tags in this scenario?

3. How do you apply the monitoring rule to the subset of hosts with the tags I mentioned above?  The documentation refers to metadata filtering but gives no useful example.  You would think they would give an example using either tags or host group.

Thanks very much for any assistance!

[StrangerThing]

We are using, and I have been told by Dynatrace to use the OS service monitoring for stuff like this. Those rules can be applied at the environment, host group, or host level. I would recommend host groups myself, as it's an easy way to group hosts and their configurations. 

[djmnyc]

Thank you.

Do you have an example of how you limit by host group?  The key value limiter text box entries are very confusing.

[StrangerThing · Accepted solution]

You would simply add the OS service monitoring policy in the host group settings instead of the global settings, so that policy would only apply to that specific host group.

[djmnyc]

Thanks very much for your feedback.  Your reply in addition to some YouTube videos I found on OS service monitoring have solved my problem.