locked
Cluster MP (6.0.6720.0) Resource Discovery Question RRS feed

  • Question

  • In this version of the the Cluster MP, the discovery of cluster resources is disabled by default.  What behavior should be expected when upgrading from Cluster MP version 6.0.6568.0?  Will discovered the relationships of clustered resources which previously existed eventually be removed?  In my pre-production environment, I've upgraded the MP and even ran the remove-disabledmonitoringobject.remove-disabledmonitoringobject cmdlet.  But, the resources remain in discovered inventory.  I may actually need to configure the override allowing discovery of cluster resources on a particular Cluster.  I do not want to create monitors on these objects until I can be sure I am controlling their discovery.  Thanks. 

    Tuesday, November 17, 2009 8:09 PM

Answers

  • Hey Stephen,

    The expected behavior is that the resources should disappear after the upgrade. The exception are the CSV resources in a hyper-V Cluster.
    Have you ensured that the discovery has run successfully after importing Cluster MP 6720? Can you please check for any errors/failures in the eventlog regarding discovery? Also, is this happening for all your clusters, or some particular cluster only?

    Can you try the following and see if you can revert to correct behavior:
    - Set DiscoverResources to true. Allow discovery to run. All resources in scope should be discovered (they already are)
    - Set DiscoverResources to false. Allow discovery to run. All resources in scope should disappear (except CSV resources).


    Regards,
    Tayyaba Rehman
    This posting is provided "AS IS" with no warranties, and confers no rights.
    Use of included script samples are subject to the terms specified at http://www.microsoft.com/info/cpyright.htm
    • Marked as answer by Marnix WolfMVP Friday, December 18, 2009 8:33 AM
    Wednesday, November 18, 2009 7:19 PM
  • If you'd like to speed things up, you can try using tasks on Cluster Service to force running the discovery manually.

    If you do end up using them, make sure you use the right one. On Win2008 clusters there are two tasks one for 2008 and one for 2008 R2. Also if you made any overrides to the discovery in the MP, you need to provide the same exact settings to the task to avoid any discovery churn.
    This posting is provided "AS IS" with no warranties, and confers no rights. Use of included script samples are subject to the terms specified at http://www.microsoft.com/info/cpyright.htm
    • Marked as answer by Marnix WolfMVP Friday, December 18, 2009 8:34 AM
    Friday, November 20, 2009 5:55 PM

All replies

  • Hey Stephen,

    The expected behavior is that the resources should disappear after the upgrade. The exception are the CSV resources in a hyper-V Cluster.
    Have you ensured that the discovery has run successfully after importing Cluster MP 6720? Can you please check for any errors/failures in the eventlog regarding discovery? Also, is this happening for all your clusters, or some particular cluster only?

    Can you try the following and see if you can revert to correct behavior:
    - Set DiscoverResources to true. Allow discovery to run. All resources in scope should be discovered (they already are)
    - Set DiscoverResources to false. Allow discovery to run. All resources in scope should disappear (except CSV resources).


    Regards,
    Tayyaba Rehman
    This posting is provided "AS IS" with no warranties, and confers no rights.
    Use of included script samples are subject to the terms specified at http://www.microsoft.com/info/cpyright.htm
    • Marked as answer by Marnix WolfMVP Friday, December 18, 2009 8:33 AM
    Wednesday, November 18, 2009 7:19 PM
  • Tayyaba,

    Thanks for the response.  After a while, the resources did start to go undiscovered.  I am seeing different behavior in my Pre-Prod than from my Production environment.  In Pre-Prod, most all the cluster resources went away with the exception of cluster "x" which is multi homed.  oddly enough, in production Cluster "x" went undiscovered.  Out of 4226 discovered cluster resources, only 239 still reside in OpsMgr.  Now, some of these can be attributed to the inability to remove retired cluster node resources / nodes from the console - a known bug.  In one case, I have an active cluster with a node down (being fixed).  And, others remain which cannot be explained.  I see no events on the cluster nodes or RMS indicating an issue but I may not know what to look for well enough.

    I will attempt the Set DiscoveredResources procedure to see if anything changes and follow up with the results.  I have a case open with premier related to this (109111349740778) if you want to review the opsmgr traces and eventlogs.

    Thursday, November 19, 2009 1:53 PM
  • If you'd like to speed things up, you can try using tasks on Cluster Service to force running the discovery manually.

    If you do end up using them, make sure you use the right one. On Win2008 clusters there are two tasks one for 2008 and one for 2008 R2. Also if you made any overrides to the discovery in the MP, you need to provide the same exact settings to the task to avoid any discovery churn.
    This posting is provided "AS IS" with no warranties, and confers no rights. Use of included script samples are subject to the terms specified at http://www.microsoft.com/info/cpyright.htm
    • Marked as answer by Marnix WolfMVP Friday, December 18, 2009 8:34 AM
    Friday, November 20, 2009 5:55 PM
  • No activity for 30 days, will mark this thread as answered now. Feel free to open it again.
    Best regards, Marnix Wolf

    (Thoughts on OpsMgr)
    Friday, December 18, 2009 8:34 AM