NOT monitored servers

Ever run into NOT monitored servers?
NOT monitored servers
NOT monitored servers
Unsure your experience, but recently, I’ve run across multiple occurrences where servers show up as ‘not monitored.  As a result, I’ve found three distinct sets of issues that might cause ‘not monitored’ status –
1) Orphaned Agent blog
This scenario usually entails deleting server from Managed Agents view in Administration tab, where alerts or other details may still exist.  The procedure leverages Holman’s orphaned agent blog (tried and true) post from years back to aid cleanup.
2) Do you have packs or connectors extending classes?
3) Rebuilding a server with the same name is a common server occurrence
Related to 1, Holman’s orphaned agent blog to be used to cleanup.
First, let’s test in the Lab.  Second, let’s talk about the express lane ‘easy button’.  Begin by deleting the Windows Computer orphaned object GUID.  Process the Windows Computer object (bottom), followed by the top two (2) are HealthService, and HealthServiceWatcher object properties (see three items highlighted).
16db02 properties
16db02 properties
Second piece, marking the Windows Computer GUID for deletion (IsDeleted=1) cleans up nearly ALL properties.  See the progress below, how this slight change makes short order of orphaned properties for ‘server’.
Windows Computer object marks all but SCVMM for deletion
Windows Computer object marks all but SCVMM for deletion
Third HealthService & SCVMM objects require manual deletion per GUID.
Note first screenshot shows health service properties marked ‘IsDeleted’ = 1 after manually processing each GUID.
HealthService marked for deletion
HealthService marked for deletion
Fourth, screenshot shows there the HealthServiceWatcher property is marked for deletion (IsDeleted=1)
HealthServiceWatcher marked for deletion
HealthServiceWatcher marked for deletion
If you have SCVMM, you will need to repeat for each of the SCVMM properties to clear out the orphans in the DB.
Why – the issue:
Typically, when servers are reimaged, i.e. NOT deleted from SCOM, there are two+ healthservice, HealthServiceWatcher, Windows Computer properties created for each image of example server.   Additional properties may show duplicated for any class discoveries that are common to the old and new image.
NOTE: Deleting the current agent may clean up objects for that instance of the discovered server, but NOT the old server image.

Verify OMS Managed Computers

Ever wondered what objects are setup for OMS?

 

Maybe you’ve seen lots of errors on servers you don’t expect ?

It’s possible someone chose a group or nearly all managed computers in your SCOM environment.

 

How do we verify, or change what computers send data to OMS from SCOM?

1) Look for a group
In SCOM console, monitoring tab

Look for the ‘advisor’ group
Maybe someone put a group in there

2) Verify OMS members

In the SCOM console, Administration tab
Click on Managed Computers
See middle pane for what is currently set up

 

Update OMS Managed computers

In the SCOM console, Administration tab
Click on Managed Computers
See middle pane for what is currently set up

Click the ‘Add a computer/group’ link on the tasks pane (right side)

Add computers or groups

Add keyword, click search, highlight and click Add

Click OK when done updating members

 

Optionally, highlight the member, click delete

 

Verify the Advisor MP’s on computer

Go to server (added or removed)

If added, look for 1201 events in the Operations Manager Log

If removed, look for 1204 events in the Operations Manager Log

 

Enjoy!!