locked
multiple test-mapiconnectivity failures against system mailbox for public folders after removing a 2007 server RRS feed

  • Question

  • After decommissioning an Exchange 2007 server (cleanly, the uninstall encountered no errors), I've started to receive alerts from SCOM that state...

    ===snip===

    Some of the MAPI connectivity transactions failed. Detailed information:

    Target: System mailbox for 2010SERVER\Public Folders

    Error: [Microsoft.Exchange.Data.Storage.MailboxUnavailableException]: Cannot open mailbox /o=MYORG/ou=Exchange Administrative Group (FYDIBOHF23SPDLT)/cn=Configuration/cn=Servers/cn=2007SERVER/cn=Microsoft System Attendant. Inner error [Microsoft.Mapi.MapiExceptionUnknownUser]: MapiExceptionUnknownUser: Unable to make connection to the server. (hr=0x80004005, ec=1003)
    ===snip===

    This same alert occurs for three different Exchange 2010 mailbox servers, in two different sites, in two different DAGs, all outside of the site where the 2007 server lived (and there are 2010 servers in that site, but they're not throwing alerts).

    When I run the Test-MAPIConnectivity command manually, it always succeeds. If I reset the health on the underlying KHI monitor, it reports a failure again within 5 minutes. I've been searching the forums for the past two days, with no luck at finding a solution. Any ideas?

    Thanks,

    Mike

    Friday, February 10, 2012 8:10 PM

Answers

  • All of the servers were rebooted this past weekend during our monthly maintenance period, and that seems to have resolved the problem. The SCOM servers had been rebooted previously, so I don't believe they were the cause of the problem. I had also restarted the System Center Health service on all of the Exchange 2010 servers (and even deleted the files in the health store folder) prior to maintenance, which also did not solve the problem. It wasn't until the Exchange servers themselves were rebooted that the issue disappeared. I'm keeping an eye out for those alerts, but it's been three days since I've seen any of those alerts in SCOM, so I am optimistic that the problem is resolved.

    Mike

    • Marked as answer by Mike Koch Monday, February 20, 2012 3:07 PM
    Monday, February 20, 2012 3:07 PM

All replies

  • Hello Mike,

    Have you read the following thread:

    Alert on transaction failures occurred during MAPI connectiving testing against an Exchange 2007 server which was removed
    http://social.technet.microsoft.com/Forums/en-US/operationsmanagergeneral/thread/7269df82-8ee5-4384-b8c0-5e5aece72b33

    It disscussed an exactly same issue. Please try the suggestions from Nicholas and Dan.

    Thanks,


    Yog Li

    TechNet Community Support

    Wednesday, February 15, 2012 10:21 AM
  • Hello Yog,

    I've read the thread you mentioned, but both seem to assume that the decommissioned server still exists in SCOM, which is not the case for us. The SCOM agent was cleanly removed, and the computer object was deleted from SCOM prior to uninstalling Exchange and removing the machine from the domain. I've reviewed the discovered inventory in SCOM, and I can find no reference to the decommissioned server anywhere. I've stopped the System Center service and cleared out the health store files on each of the servers that are generating the alerts, but they come back shortly after restarting the service.

    This weekend is our monthly maintenance, so all of the servers will be rebooted after patching. I'm hoping that restarting the SCOM servers, including the correlation engine, will resolve this issue. Otherwise, I may have to dig into it a bit more.

    Thank you for your help. I appreciate it.

    Mike

    Wednesday, February 15, 2012 9:03 PM
  • All of the servers were rebooted this past weekend during our monthly maintenance period, and that seems to have resolved the problem. The SCOM servers had been rebooted previously, so I don't believe they were the cause of the problem. I had also restarted the System Center Health service on all of the Exchange 2010 servers (and even deleted the files in the health store folder) prior to maintenance, which also did not solve the problem. It wasn't until the Exchange servers themselves were rebooted that the issue disappeared. I'm keeping an eye out for those alerts, but it's been three days since I've seen any of those alerts in SCOM, so I am optimistic that the problem is resolved.

    Mike

    • Marked as answer by Mike Koch Monday, February 20, 2012 3:07 PM
    Monday, February 20, 2012 3:07 PM