SCVMM 2012 R2 – Error 20506 / 20552 / 2606

Had an interesting issue today, All hosts spread over two clusters were in a state of ‘not responding’ with no clear explanation why.

It looked as though the issue began at 2am, a time when nobody would have been working on the VMM console.

Initial diagnosis showed that the network was up, The Failover Cluster was working just fine, but VMM had lost WinRM connectivity to the hosts.

The usual error 20506 appeared in the host status window indicating that WinRM “could not complete the Windows Remote Management request on computer xxx”

Furthermore each time the Refresh Host Cluster job would fail it would show errors

2606 – unable to perform the job because on or more of the selected objects are locked by another job

20552 – VMM does not have appropriate permissions to access the resource on the xxx server

I spent a fail amount of time looking into the WinRM issue, everything was fine and nothing had changed. WinRM tests showed that the VMM server was able to communicate with the hosts.

I looked for locked jobs or running jobs, even checking the SQL server using the query

SELECT * FROM [VirtualManagerDB].[dbo].[tbl_VMM_Lock] where TaskID=’Task_GUID’ (See http://support.microsoft.com/kb/2795040 )

but that showed no jobs were active at all.

 

The thing that resolved it, was running another job (a non cluster related job) in my case i changed a run-as account. As soon as this went through I was able to refresh the cluster just fine.

I cant completely explain this one, my guess is that a phantom job had gotten stuck somewhere. Thought i would share the results.

 

9 Responses to SCVMM 2012 R2 – Error 20506 / 20552 / 2606

  1. Karl says:

    This also happened to me today. Thanks!

  2. Kirpal Singh says:

    Hi

    I have Shifted Standalone VMM 2012 Sp1 in HA mode.
    Now, I faced issue that is Host not responding with event 20552.
    if Service is running is node1, Everything is ok.
    if Service is shifted to node 2 from node1 then there is issue host not responding, if I set runas account password , then it is working fine on Node2.
    if I again shift service to node1 then run as account need to be set again in node1, Kindly help

  3. Yoni says:

    Legend!

    you saved my bacon.

    Thanks

  4. kirpal singh says:

    We solved it, MS Team Suggested me to Remove its DB from SQL Availability Group, Remove SCVMM from passive node and Active node with retain database and Reinstall it with Existing database and its database again to SQL Availability Group

  5. Neil Levien says:

    I had the same issue after having to reinstall. All I ended up doing was revalidating the credentials of my run as account.

    • alex says:

      I got the same issie. I have the following setup:

      -hyper-v cluster running at the datacenter
      -vmm-server with APP-controller running on a different location.

      The vmm server is up and running and I’m trying to add the 1 hyper-v server from the datacenter. I installed it as “windows server computer in a perimeter network”. I followed all the right steps from this page (only i took the install files from the vmm-server, not from the cd)

      http://www.windows-noob.com/forums/index.php?/topic/11270-how-to-add-hyper-v-hosts-in-a-perimeter-network-in-vmm-2012-r2/

      When I’m trying to add the hyper-v server, I got the error following error:

      Error (20552)
      VMM does not have appropriate permissions to access the resource on the 46.21.156.34 server.

      Recommended Action
      Ensure that Virtual Machine Manager has the appropriate rights to perform this action.

      Also, verify that CredSSP authentication is currently enabled on the service configuration of the target computer 46.21.156.34. To enable the CredSSP on the service configuration of the target computer, run the following command from an elevated command line: winrm set winrm/config/service/auth @{CredSSP=”true”}

      If someone knows how to fix it, i would really appreciate it

      • I’m not 100% but I would try using a hostname/DNS to resolve the computer rather than just the IP. Also that IP is a public IP address by the looks, so you shouldnt be communicating over that channel.

  6. JuuGo says:

    Just Reset your run as account password or use a Service Account 🙂

Leave a reply to Kirpal Singh Cancel reply