Exchange Server 2003 / Exchange HTTP Virtual Server

Discussion in 'Windows Server Systems' started by madmatt, Jul 21, 2005.

  1. madmatt

    madmatt Bow Down to the King Political User

    Messages:
    13,312
    Location:
    New York
    I am running a two node cluster for Exchange Server 2003.

    I tried to access OWA tonight and the site timed out. So I logged into the server to check the event log. I see the following error:

    "Exchange HTTP Virtual Server Instance 100 (GRENO-EXCHANGE): Failed to get the protocol IP address and port bindings from the metabase."

    I open Cluster Administrator and try to bring the resource online but it fails and provides the same event error as above.

    Firstly, what would have caused this?

    Secondly, how do I resolve this issue?

    Any help is appreciated. Thank you.
     
  2. kcnychief

    kcnychief █▄█ ▀█▄ █ Political User Folding Team

    Messages:
    16,948
    Location:
    Massachusetts
    I know this article is for Windows 2000...

    http://support.microsoft.com/kb/838922

    But it looks like the information is still relevant.....maybe?

    I don't know much about your setup, but it makes sense.
     
  3. madmatt

    madmatt Bow Down to the King Political User

    Messages:
    13,312
    Location:
    New York
    No headers have been added. I've found documentation on Microsoft's newsgroups saying the metabase.bin file may have become corrupt. Which doesn't sound like a fun fix.

    Any other ideas?
     
  4. madmatt

    madmatt Bow Down to the King Political User

    Messages:
    13,312
    Location:
    New York
    The problem has been resolved.

    In the process of configuring the HTTP protocol through Exchange System Manager it made changes to the MetaBase XML file (configuration). Some how the "ServerBindings" property was removed from Exchange Virtual Server configuration. To resolve the issue I had to shutdown IIS on the inactive node and edit the file in a plain text editor and add the "ServerBindings" property back in with the correct information. After which I restarted the server and then failed over the cluster to this node.

    Making the change on one server affected both so no further work was needed.

    I hope this helps someone else if they run into this problem.
     
  5. kcnychief

    kcnychief █▄█ ▀█▄ █ Political User Folding Team

    Messages:
    16,948
    Location:
    Massachusetts
    Interesting fix, it appears they replicated to eachother which in essence caused the problem and at the same time, fixed it :)
     
  6. madmatt

    madmatt Bow Down to the King Political User

    Messages:
    13,312
    Location:
    New York
    It makes sense that the replicated the configuration since the web site is a virtual server.

    It took some digging and I just got too tired dealing with it last night. But this morning I got on the right track with the XML file (a 1000+ line XML file is a lot of fun to dig through looking for discrepancies).
     
  7. kcnychief

    kcnychief █▄█ ▀█▄ █ Political User Folding Team

    Messages:
    16,948
    Location:
    Massachusetts

    Sometimes it is best to leave it and come back with a fresh head. You are lucky your the admin, in some places i restrict the F3 search feature. I bet that comes in handy for 1,000 lines
     
  8. madmatt

    madmatt Bow Down to the King Political User

    Messages:
    13,312
    Location:
    New York
    Ha, Ha. Yes. F3 is my friend.
     
  9. kcnychief

    kcnychief █▄█ ▀█▄ █ Political User Folding Team

    Messages:
    16,948
    Location:
    Massachusetts
    Me personally, I like to place F3 on the pillow next to me and cuddle with it at night :laugh:

    *has issues*
     
  10. madmatt

    madmatt Bow Down to the King Political User

    Messages:
    13,312
    Location:
    New York
    Don't we all (have issues).
     
  11. Bootsy

    Bootsy Huh?

    Messages:
    1,124
    Location:
    Miami, Fl
    Interesting problem, glad you found a fix for it, I was going to suggest to look into the connections...I was way off hehe.

    (lovin F3 as well)
     
  12. kcnychief

    kcnychief █▄█ ▀█▄ █ Political User Folding Team

    Messages:
    16,948
    Location:
    Massachusetts
    We should have a BBQ in the honor of F3 :)
     
  13. madmatt

    madmatt Bow Down to the King Political User

    Messages:
    13,312
    Location:
    New York
    For anyone who was following this thread.

    I was wrong. The XML config file did not replicate to both nodes. Once it failed back over to the first node the settings were lost. I had to make the changes to both XML configs for it to work correctly.
     
  14. markrich

    markrich OSNN One Post Wonder

    Messages:
    1
    We have begun to experinence this problem also. I am unsure what to look for in metabase.xml but all seems well. The correct IP addresses and other references to servers are all in there. In our case the HTTP Virtual Server Instance 100 refuses to load on the second node causing the services to all bounce back again.

    How did manage to fix this problem. It's driving me insane.

    Marky
     
  15. madmatt

    madmatt Bow Down to the King Political User

    Messages:
    13,312
    Location:
    New York
    You want to make sure the IP address (to the virtual server) and port 443 are bound to the Exchange Virtual Server in the IIS manager.

    If it is then I will need more information, such as event errors (event viewer).