none
Exchange 2013 ECP and DAG issue RRS feed

  • Question

  • Good Day,

    We have Exchange 2013 setup in Coexistence with Exchange 2007. Everything is working well except for two small issues I have noticed.

    Current Setup of 2013:

    SiteA:

    2x CAS Servers on Windows 2012 R2 configured with windows NLB (Internet Facing).

    2x MBX servers Configured in a DAG also running on Windows 2012 R2.

    SiteB:

    2x CAS Server on Windows 2012 R2.

    2x MBX servers Configured in a DAG also running on Windows 2012 R2.

    Issues I am having:

    Issue 1:

    When working on the ECP:

    I can do all tasks on servers in the same site but get some errors when trying to do tasks on servers in the other site.

    When trying to Activate a DB copy on the DAG at SiteB I get "The Microsoft Exchange Replication service does not appear to be running on Server name"

    And just trying to go to the properties of a server at SiteB "An error occurred while accessing the registry on the server servername"

    This happens vice versa.

    I can access all servers on the same site but not between sites with some tasks.

    I have made sure firewall is disabled and remote registry is indeed running but no luck.

    From a network perspective all ports are open between the sites.

    Issue 2:

    When I reboot a DAG member with the active mailbox databases for testing outlook connectivity I get two issues.

    Then I cannot access ECP at all until the server is up again.

    Outlook keeps staying on trying to connect until it is reopened.

    Any Ideas?

    I will gladly give more information if needed.

    Thank You

    Riaan Rourke

    Senior System Engineer



    Monday, September 15, 2014 1:40 PM

Answers

  • Hi ,

    For issue 1:

    Make sure all the exchange servers are installed with same CU'S 

    For issue 2 :

    From your configuration i  came to knew that you are having two node DAG extended across two sites .

    What about the file share witness ? where you have kept that  ? whether it would be in site A or Site B?

    Same time in your siteA  you are having CAS servers in NLB cluster .So for cluster name there would be an host A record in DNS , So during the database failover to Site B you should have to play with DNS.

    That means just rename the HOST A record for nlb cluster name to the ip address of the cas server in SITE B .

    PLease review the event logs in all the exchange servers. 

    Please reply me if you have any queries.

    Regards

    S.Nithyanandham


    Thanks S.Nithyanandham

    • Marked as answer by Belinda Ma Friday, September 26, 2014 8:00 AM
    Wednesday, September 17, 2014 6:47 AM

All replies

  • Are these VMs?  If so what are they running on?

    Cheers,

    Rhoderick

    Microsoft Senior Exchange PFE

    Blog: http://blogs.technet.com/rmilne  Twitter:   LinkedIn:   Facebook:   XING:

    Note: Posts are provided “AS IS” without warranty of any kind, either expressed or implied, including but not limited to the implied warranties of merchantability and/or fitness for a particular purpose.

    Tuesday, September 16, 2014 3:35 AM
  • Hi Roderick,

    Yes they run on VMware,

    Regards

    Tuesday, September 16, 2014 6:32 AM
  • I have setup the whole environment according to Virtualization best practises.

    Might there be something is missed?

    Kind regards

    Tuesday, September 16, 2014 6:34 AM
  • Hi,

    For the first issue, please make sure there is no network connectivity issues.

    For the second issue, when you reboot one DAG members, auto-failover of database should start. Based on your description, auto-failover didn’t work.

    Please use the Test-ReplicationHealth cmdlet to test the replication status.

    Besides, please check the application log to see if there is any related error messages.

    Best regards,


    Belinda Ma
    TechNet Community Support


    • Edited by Belinda Ma Tuesday, September 16, 2014 9:22 AM
    Tuesday, September 16, 2014 9:20 AM
  • Hi Belinda,

    There are no connectivity issues.

    The auto-failover does work,

    The DBs goes active on the passive DAG member.

    Thank You

    Tuesday, September 16, 2014 1:08 PM
  • Since I don't see this with Hyper-V I'll assume this is on ESX. Please let me know it that's not correct.

    In terms of virtualization best practices - whose list was that :) 

    There are multiple configuration things that you *MUST* do when using WLBS on ESX.  It will depend upon multiple things in the environment.  I'd start with this

    http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1556

    Fix that, and chances are the other item will be also remediated.


    Cheers,

    Rhoderick

    Microsoft Senior Exchange PFE

    Blog: http://blogs.technet.com/rmilne  Twitter:   LinkedIn:   Facebook:   XING:

    Note: Posts are provided “AS IS” without warranty of any kind, either expressed or implied, including but not limited to the implied warranties of merchantability and/or fitness for a particular purpose.

    Tuesday, September 16, 2014 2:22 PM
  • NLB in Unicast was an issue initially but I have fixed that.

    Currently running in Multicast mode.

    And the issue between the sites any ideas?

    Thanks

    Wednesday, September 17, 2014 6:06 AM
  • Hi ,

    For issue 1:

    Make sure all the exchange servers are installed with same CU'S 

    For issue 2 :

    From your configuration i  came to knew that you are having two node DAG extended across two sites .

    What about the file share witness ? where you have kept that  ? whether it would be in site A or Site B?

    Same time in your siteA  you are having CAS servers in NLB cluster .So for cluster name there would be an host A record in DNS , So during the database failover to Site B you should have to play with DNS.

    That means just rename the HOST A record for nlb cluster name to the ip address of the cas server in SITE B .

    PLease review the event logs in all the exchange servers. 

    Please reply me if you have any queries.

    Regards

    S.Nithyanandham


    Thanks S.Nithyanandham

    • Marked as answer by Belinda Ma Friday, September 26, 2014 8:00 AM
    Wednesday, September 17, 2014 6:47 AM