none
Issues backing up primary DPM server with 2ndary DPM server - DPM 2010 RRS feed

  • Question

  • I am trying to backup my primary DPM server using a secondary DPM server.  Both DPM servers are using DPM 2010 on Server 2008 R2.  If I select a replica to protect from the protection group creation on the 2ndary server - at the part where it provisions the disk space - there is an error about unable to communicate with 'Protected Server'.  I am reading this page : http://technet.microsoft.com/en-us/library/bb795680.aspx and towards the bottom - it tells me to do the following:

    "

    To enable protection of the replicas, you must establish communications between the secondary DPM server and the data sources protected by the primary DPM server. When local data sources are protected on a cluster node, you should enable protection for each cluster node in addition to the cluster. To establish communications, perform the following steps:

    1. On the Agents tab in the Management task area, select a primary DPM server.
    2. In the Details pane, next to Protected servers enabled for protection, click Modify."

    I have no details pane that I can see nor is there a Modify button.  Can anyone help me please?

    Thanks,

    Zaach


    Zach Smith
    Tuesday, February 8, 2011 11:11 PM

Answers

  • OK - so I found the 'real' solution to this issue - here it is if anyone is interested.

    I was just backing up the dpm volume C:\program files\dpm\volumes... whatever as a work around.  I'd just find the volume for the server I wanted and back that up from my 2ndary DPM server.  So here is what I did to fix it so i can back it up 'normally'. 

    It occurred to me that this host machine (my server that i oculdn't back up from my 2ndary dpm serveR) has two nics each in two different subnets.  So I  did a tracert -d 2ndarydpm and the results were the interface of my dpm server for all hops - very odd.  So i went into the 'route print' of the virtual host server and removed the 0.0.0.0 0.0.0.0 route that used the interface in my other subnet as I did not need this and that caused my tracert -d 2ndarydpm server to be correct and show each router for each hop instead of its local interface for each hop.  This then allowed my 2ndary dpm server to communicate and allow the host to be backed up normally.

    Does that make sense?

    Zach


    Zach Smith
    • Marked as answer by ZachSmith Thursday, December 22, 2011 3:45 PM
    Thursday, December 22, 2011 3:45 PM

All replies

  • Hi Zach,

    The article you linked is for DPM 2007. The secondary protection technet article for DPM 2010 is different - http://technet.microsoft.com/en-us/library/ff634186.aspx 

    It sounds like the DPM server and the DP agent on the protected server are not able to communicate. You will need to get the exact error message to troubleshoot. Once you have that you can start here http://technet.microsoft.com/en-us/library/ff399208.aspx

    If you still can't get it work, post back here with the exact error message.

    Thanks,

    Marc

     

    Wednesday, February 9, 2011 2:18 PM
  • Thanks for the updates.  I will read through these sites.  If you are interested - below is the error message.  Again - I see this error message After I select to backup a hyper-v child snapshot and I get to the part where it calculates the disk needed for the backup.

    "DPM Cannot browse "Protected-Server.domain.com" because the agent is not responding.  If you just installed an agent on "protected-server.domain.com", the computer might be restarting.  Wait a few minutes after Windows starts for the agent to become available.  Otherwise, troubleshoot the problem as follows:

    1)Check recent records from the DPMRA source in the application event log on 'protected-server.domain.com' to find out why the agent failed to respond.

    2)Make sure that the DPM server is remotely accessible from 'protected-server.domain.com;

    3)If a firewall is enabled on the DPM server, make sure that it is not blocking requests from 'protected-server.domain.com'

    4)Restart the DPM protection agent service on 'protected-server.domain.com'  If the service fails to start, re-install the protection agent.

    ID: 33

    Details: Internal error code 0x8099090E"

     

    Protection works great from the primary server - but the secondary server gets this message.

    Zach


    Zach Smith
    Wednesday, February 9, 2011 11:23 PM
  • Have you tried this when there is no backup running on primary for protected server.

     Are protected-server and the secondary DPM in same Domain?

    Can you please check the Event viewer of Protected-Server and Secondary server at the time this happened.

    It might give us some idea of what is going wrong.

     


    Thanks, NikhilKumar.R [MSFT] - This posting is provided "AS IS" with no warranties, and confers no rights.
    Thursday, February 10, 2011 1:48 PM
  • I am getting this error message on the server i'm trying to protect.

    "DCOM was unable to communicate with the computer 'secondar-DPM-server.domain.com' using any of the configured protocols."

    The primary and secondary DPM server are members of the Distributed COM users group on the protected server.  I am investigating this.  One thing to note - this message pops up when I am creating the protection group and I click the + sign next to the protected server under the primary's protected servers list.  When I click the + on the specific server - it takes a couple of minutes to exapnd and display 'HyperV' 

     

    I'll keep you posted on anything I figure out.  Thanks for the help.

    Zach


    Zach Smith
    Thursday, February 10, 2011 3:58 PM
  • On the server i'm trying to protect from my 2ndary DPM server - I do notice tthe below as well.

    System Event:

    The WMI Performance Adapter service entered the running state.

    Application Event:

    Windows Management Instrumentation has stopped WMIPRVSE.EXE because a quote reached a warning value.  Quotea: HandleCount Value: 6746.  Maximum value: 4096 WMIPRVSE PID:7240

    Not sure if this is related to my issue yet.  Just throwing it out there

    Zach


    Zach Smith
    Thursday, February 10, 2011 4:26 PM
  • I was able to remove the Application Event - Windows Management Instrumentation has stopped WMIPRVSE.EXE because a quota reached a warning value... by the following the instructions from http://social.technet.microsoft.com/Forums/en/virtualmachinemanager/thread/e9e704dc-bfab-498b-8d9c-40c9d390b73c

    I still get the DCOM issues though.  I noticed that on the protected server - in the System Event logs - when I try to expand the protected server from the protectiong group wizard from the 2ndary server - I get "The WMI Performance Adapter service entered the stopped state" and "The WMI Performance Adapter entered the running state" - Then I get 4 DCOM errors ""DCOM was unable to communicate with the computer 'secondar-DPM-server.domain.com' using any of the configured protocols." then the WMI Performance Adapter starts/stops - then 4 more DCOM errors - then the WMI Performance adapter stops. 

    Then in the protect group wizard - The protected server is now exapnded and shows 'HyperV'  I select what I want to backup and when I get to the calculating disk space part - The DCOM errors come again then the protection group wizard fails.

    On the protected servers - the 2ndary and Primary DPM server are in the DCOM users group + the 2 DPM groups.

    Any help is greatly appreciated.

    Zach


    Zach Smith
    Thursday, February 10, 2011 5:15 PM
  • I tried to backup a hyperV instance from our HyperV cluster (the server in question for this post is a stand alone hyperV host) and the backup of the cluster works successfully.  Perhaps I need to reinstall the agent on the stand alone hyperV server.  Not sure what is going on but since it works on other machines that seems to be a good place to start.  This machine has a lot of important stuff on it so I will do this at night - most likely over the weekend.  I'll report back when I reinstall the agent.  If anyone has any other ideas in the mean time please let me know.

    Zach


    Zach Smith
    Thursday, February 10, 2011 7:16 PM
  • Zach,

    can you please report us your status?

    sorry for the delay but my reply could be out of date for above things so waiting for present update so that

    i can reply on top of that.


    Thanks, NikhilKumar.R [MSFT] - This posting is provided "AS IS" with no warranties, and confers no rights.
    Monday, May 16, 2011 12:36 PM
  • Nikhil,

    Sorry for the late response.  I am still having this issue although I have implemented a workaround.  I doubt this is a recommended or even supported method but it is working.  The issue is my 2ndary DPM server cannot backup VMs from ONE virtual host machine that my primary DPM server is protecting.  I get DCOM errors.  My primary can back this server up, my 2ndary can back up other protected volumes from this primary server, but it just cannot back up any VMs from one particular virtual host machine.  I've checked DCOM users group and permissions look correct. 

    Here is what I did.

    From the 2ndary DPM server instead of browsing to Primary DPM - Protected Servers - Virtual Host - HyperV - VMs - I determined where the VMs (.vhd files really) ARe stored in the DPM storage pool - C:\Program Files\Microsoft DPM\DPM\Volumes\... and I am backing up my VMs from this one host this way.  It works but like I said - probably not recommended or supported.

    Thanks,

    Zach


    Zach Smith
    • Marked as answer by ZachSmith Monday, November 21, 2011 5:32 PM
    • Unmarked as answer by ZachSmith Thursday, December 22, 2011 3:41 PM
    Monday, November 21, 2011 5:32 PM
  • One thing to add to my marked-answer post - I'm not sure why 2ndary DPM servers dont 'logically' do what I did as a workaround.  Why would my 2ndary DPM server need to start the Remote Agent on the host machine it's backing up?  Why wouldn't I essentially be backing up the DPM storage pool.  My primary server reacts with the Remote AGent on the protected server and a secondary server backing that up should only communicate with the primary DPM server and it's storage pool.  I would think it would make more sense to only have the 2ndary server do this instead of having it also communicate with the machine being backed up.  What I am doing is essentially that and it works.

    That could end up to be a lenghty conversation but just thought i'd throw it out there.

    Zach


    Zach Smith
    • Marked as answer by ZachSmith Thursday, December 22, 2011 3:45 PM
    • Unmarked as answer by ZachSmith Thursday, December 22, 2011 3:45 PM
    Monday, November 21, 2011 7:29 PM
  • OK - so I found the 'real' solution to this issue - here it is if anyone is interested.

    I was just backing up the dpm volume C:\program files\dpm\volumes... whatever as a work around.  I'd just find the volume for the server I wanted and back that up from my 2ndary DPM server.  So here is what I did to fix it so i can back it up 'normally'. 

    It occurred to me that this host machine (my server that i oculdn't back up from my 2ndary dpm serveR) has two nics each in two different subnets.  So I  did a tracert -d 2ndarydpm and the results were the interface of my dpm server for all hops - very odd.  So i went into the 'route print' of the virtual host server and removed the 0.0.0.0 0.0.0.0 route that used the interface in my other subnet as I did not need this and that caused my tracert -d 2ndarydpm server to be correct and show each router for each hop instead of its local interface for each hop.  This then allowed my 2ndary dpm server to communicate and allow the host to be backed up normally.

    Does that make sense?

    Zach


    Zach Smith
    • Marked as answer by ZachSmith Thursday, December 22, 2011 3:45 PM
    Thursday, December 22, 2011 3:45 PM