locked
A CCR node is not monitored for its replication health state RRS feed

  • Question

  • Hi all,

    here is the issue, i got a CCR, where one node is monitred and the other one isnt for "replication health state", the node is in scom but its state is "Not monitored". Is it posible that i manualy turned off the monitoring on this node? How can i check if i did this. If this is the problem which im not sure but it could be, how can i turn in back on?

    Its SCOM 2007 R2 MP for Exchange

    Tnx in advance! 

    Wednesday, July 7, 2010 8:36 AM

Answers

  • Not monitored usually is driven by one of two things:  Either the discovery is not working (and thus you don't see the nodes in the state view at all) or the workflows for those nodes are either disabled or failing to run.  Look in the agent log on these servers to see if there are errors in the operaitons manager event log.  Also, look at the failed workflow report for those agents under the agent health on the effected computers.  If there are multiple monitors, make sure someone hasn't disabled them specifically with an override on a group that excludes the non-monitored computers.
    Microsoft Corporation
    • Marked as answer by Yog Li Friday, July 23, 2010 11:20 AM
    Saturday, July 10, 2010 3:14 PM

All replies

  • If CCR stands for fail over cluster (MSCS) then it is normal that only the active node in a cluster is monitored.  You cannot change it.
    Microsoft Corporation
    Wednesday, July 7, 2010 2:46 PM
  • Hi Dan,

    Its localed in Microsoft Exhcange Server 2007 > Server State > CCR node (MBX2) , and its state is not monitored, for the other CCR that we have both nodes are monitored. Non of the health monitores are working except of Tes treplication health on the not monitored node.

    Thursday, July 8, 2010 8:01 AM
  • Not monitored usually is driven by one of two things:  Either the discovery is not working (and thus you don't see the nodes in the state view at all) or the workflows for those nodes are either disabled or failing to run.  Look in the agent log on these servers to see if there are errors in the operaitons manager event log.  Also, look at the failed workflow report for those agents under the agent health on the effected computers.  If there are multiple monitors, make sure someone hasn't disabled them specifically with an override on a group that excludes the non-monitored computers.
    Microsoft Corporation
    • Marked as answer by Yog Li Friday, July 23, 2010 11:20 AM
    Saturday, July 10, 2010 3:14 PM
  • Tnx Dan, ill check that and report back
    Saturday, July 10, 2010 4:38 PM
  • Hi,

    There are no errors in the logs, and there are no overides for the monitors that are on the problematic server. The Healt of the server it self is Healty, but the Exhchange server state is Not Monitored. The Availabitlty monitor (non of the disks that conatins the databases and the logs are monitored), and the Configuration monitor are not working (only a white circle)

    Monday, July 12, 2010 7:14 AM
  • Hi, as you state, there should be disk and log shipping monitors that apply to the CCRNode via dependency monitors. There are no monitors that apply directly to the CCRNode. Do you see the expected disks and log shipping instances discovered for the server? You can see them for example by opening the diagram view for a server (right click on the server role).
    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
    Wednesday, July 21, 2010 1:02 AM
  • Hi Ake,

    I see all of the disks for log shipping, but they arent showing as monitored. As far as i checked their are no overides for this specific node.

    All the best,

    Zarko

    Wednesday, July 21, 2010 8:49 AM