locked
Exchange 2010 SP1 ClientAccessServer freezes after event 5783 RRS feed

  • Question

  • Hi everyone, we are running a simple exchange 2010 SP1 installation in a domain with 3 domain controllers (windows server 2008R2 SP1, domain level Windows Server 2008 R2), an exchange mailbox and an exchange client access server, both virtual machines running 2008R2 SP1 as well).

    Our client access server froze several times within the last few month (it was not accessible from network, cpu-utilization in Hyper-V-Manager fluctuated between 45 and 50%). Only a hard reset could put it back into service.

    In most cases no event was logged, but last week we got system event 5783:

    Event ID: 5783: The session setup to the Windows NT or Windows 2000 Domain Controller \\CAS1 for the domain *** is not responsive.  The current RPC call from Netlogon on \\the-computer to \\************.com has been canceled.

    Also application event 2128 (object xxx not found because of replication or security problems?!) is frequently logged at both exchange servers.

    Running dcdiag /v did not show any problem and repadmin /showrepl ran successfully.

    Any ideas what's going wrong?

    Thanks in advance, Peter

    Tuesday, March 15, 2011 8:06 AM

All replies

  • Hi
       Event 2128 is only warning event. It won’t cause any error.  
       You should run adsiedit connect to your dc to check the  object.
       If you could see the object when connecting to other DC, it can be caused by a replication or permission issue.  If so, check the permission of that object, and compare with other value.
      
    Event 5183
       You should check the network and status of your DC.
        If the source server could not locate the server in DNS, troubleshoot Active Directory replication failure due to incorrect DNS configuration. If this is not a DNS problem, troubleshoot RPC problems.
        All the these two error can tell me that there is something wrong of network or your DC.
        Can you run Exbpa to check your CAS server again ?
    Wednesday, March 16, 2011 3:46 AM
  • Hi Terence, thanks for your answer!
    To Event 2128:
    In adsiedit we can find all the objects claimed by the warnings, on all 3 DCs, from both exchange servers?!
    To Event 5183:
    We recently migrated our Domain controlles from SYSVOL to DFS Replication. We detected no replication or permission issues until now. Running the replication diagnose-report on Domain System Volume doesn't show any errors ... while our CAS-box frooze before and after the migration.
    The only issues we detected are DNS errors: Running dcdiag /test:DSN results in a couple of
    - "Missing AAAA-Records" in gc._msdcs.domain,
    - "Missing SRV-Records" in _ldap._tcp.domain and _kerberos._tcp.dc._msdcs.domain on all 3 DCs, which are DNS-servers also.

    All the claimed zones do have valid IPv4 entries and we can see all the SRV-Records claimed to be missing?!

    If this (IPv6) could be the reason for our freezing CAS, is there a way to "automatically register" the missing IPv6 entries? ipconfig /registerdns doesn't do this job ...

    Remark: We have turned off IPv6 on Mailbox- and CAS-Server via HKEY_LOCAL_MACHINE\SYSTEM\CCS\services\TCPIP6\Parameters\DisabledComponents=ffffffff, but not on our Domain Controllers 

    Thanks in advance, Peter


    Peter Seethaler

    Friday, March 18, 2011 8:11 AM