OCS (UM?) Dropping Incoming Calls after 35-40 Seconds RRS feed

  • Question

  • Hello!


    I'm having a problem with OCS (UM?) that is very recent - only a few days old.  Currently, it appears that OCS is dropping incoming calls after 35-40 seconds.  This is about enough time for someone to call in to the AutoAttendant, say a name, get transferred and have the recipient say "hi".  This issue began occurring suddenly last Friday.  Restarting the services and dredging through the configs have given me no insight into the problem, and no errors are reported in the event viewer on all systems, including the OCS pool, the mediation server, nor the edge server.  Interestingly enough, this issue seemed to have resolved itself on Friday evening, only to return again on Monday.


    A packet trace on the mediation server clearly shows the mediation server sending a RTCP "Goodbye" and a SIP "BYE" to the PBX, making me pretty certain this issue is occurring somewhere within OCS.


    I had not installed any patches nor made any configuration changes between the time that it was working and failed, then began working again, aside from reboots.  The systems did not start working again after they rebooted.


    Here's my breakdown of what does/doesn't work:

    • Communicator to Communicator - Works
    • Communicator to External Number - Works
    • External Number to Communicator - Disconnects after 35-40 seconds
    • External Number to Communicator Which is Forwarded - Disconnects after 35-40 seconds
    • External Number to Communicator Through AutoAttendant - Disconnects after 35-40 seconds

    In all cases, if we wait for the call to go to voicemail, it gets disconnected before voicemail answers.


    Does anyone have any ideas on how to dig deeper into this issue?
    • Moved by Matt Sousa - MSFT Thursday, December 17, 2009 6:59 PM forum migration (From:Telephony)
    • Moved by Matt Sousa - MSFT Thursday, December 17, 2009 7:32 PM forum migration (From:Enterprise Voice and Telephony)
    Monday, August 25, 2008 10:29 PM

All replies

  • Additional information: the performance counters on the mediation server show X number of attempted calls, but 0 successfully established.
    Monday, August 25, 2008 10:52 PM
  • I have seen such behaviour when incomung INVITE included ";user=phone". Can you do a Wireshark trace and check what's happening SIP-wise? Do you get an OK from Mediation server?



    Tuesday, August 26, 2008 11:21 AM
  • At the moment, it's working fine again.  The incoming invites definitely have ;user=phone, even now, while it's working.  Next time it starts failing, I'll run another pcap and make sure we're seeing an OK from the mediation server.


    Tuesday, August 26, 2008 6:13 PM
  • Ok, it's happening again.  A sip trace shows that I am receiving an OK from the Mediation Server, and all seems to be well for a few seconds, then suddenly, I get a BYE from the Mediation Server.


    Edit: There is no ;user=phone present in the SIP communications between my endpoint trunk and the Mediation server, however the Mediation Server debug logs do show that the From and To fields both have ;user=phone after the SIP URI.

    Friday, September 5, 2008 8:46 PM
  • Is there any firewall agent been blocked the connection. Try to disable the firewall agent.
    Thursday, December 10, 2009 8:51 AM