Determining what is going on with GSM using OpsMgr event collection and views (#SCOM, #SYSCTR)

The new GSM monitoring functionality in Operations Manager 2012 SP1 is really interesting, but when something goes wrong it’s difficult to troubleshoot what is occurring. When you configure your subscription for GSM, you specify a resource pool which will perform this function (as shown below). This means that any member of the pool can be providing these functions (which is good for redundancy).

clip_image005

As Daniele pointed out in his blog post (http://nocentdocent.wordpress.com/2013/02/07/global-service-monitoring-moving-service-testing-to-the-cloud/) GSM creates events in the Operations Manager log from an event source of “Health Service Modules Ex” but these events can be occurring on any member of the pool chosen when configuring GSM. An example event is shown below:

clip_image001

In my environment, the existing GSM test I had in place was providing an event approximately every 25 hours as shown below. While it is pretty easy to find these events in an environment with a single management server, this gets far more difficult when you have a significant number of management servers.

clip_image003

To provide a simple way to debug what is occurring with GSM, I created a management pack which collects GSM event information and includes a view to display these events.

The rule name is “Collect GSM Event Information” as shown below and it is targeted at the

image

The event view is shown below and shows all events collected by the rule created above.

clip_image004

This rule gathers both TFS and GSM events, but the volume was pretty small and the description field wasn’t always consistent so I decided to leave it as-is shown below. It’s also targeted at all management servers so it may be checking on systems which are not using GSM depending on how you configured your subscription.

image

This management pack is available for download at SystemCenterCenter at: http://www.systemcentercentral.com/pack-catalog/global-service-monitoring-debug-management-pack/

UPDATE:

Daniel Savage with the product team let me know that there is also an existing set of views which provide the health for the GSM components (not the events, but the health state and alerts) which are available in the Operations Manager / Global Service Monitor folder. A sample of this view is shown below! (Thank you Daniel!)

GSM Monitor state

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.