Announcement

Collapse
No announcement yet.
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Shoreware TMS disconnection issue Event id 233

    Hi All

    I currently have an issue where all ShoreTel switches lose connection to TMS about every 5 minutes or so (for a couple of seconds).

    I really want to get to the bottom of the cause, as myself and our support partners are unsure.

    We are on HQ v13.2 and ECCv8.0

    We also run all machines as VM's on the VMware platform and use Dell powerconnect switches for our edge and core.

    If anyone has a powerconnect QoS config for shoretel, I think that would be a good start, to confirm we our switching is correctly configured. unfortunately all the Shoretel recommended documentation I have been able to find is for cisco switching which means I can't check.

    If you need any more information please let me know

    Many thanks

    Graeme

  • #2
    Hi,

    Has it started suddenly?

    Is Spanning Tree running on the ports connected to ShoreTel kit? Disable this if so on those ports. Are you running teams to the HQ server? This is not supported, and I have seen problems with this in the past.

    Are there any other indicators of a network issue? Spanning Tree recalculations will cause TMS to disconnect and reconnect (you will see lots of 233 and 234 issues)

    Have you recently upgraded?

    Comment


    • #3
      The issue has been happening for a number of weeks, but we have recently upgraded (this weekend from v12.2/v7) to see if helped prevent the issue.
      The rest of the network is stable, and we have tried to configure Qos on the edge switching - but haven't had a chance with the core.

      RSTP is configured on all ports – what should they be disabled on?

      Also what do you mean by running teams to the HQ?

      Thanks for the reply!
      Last edited by gr1; 04-18-2013, 07:28 AM.

      Comment


      • #4
        First check the following on the network switch:
        - Spanning tree must be off
        - ARP table should not show duplicates/wrong information
        - If you are using the MAC security feature on a Cisco switch, there was a KB article published that mentions that causes problems (I can confirm that it does)

        Also make sure that you have a dedicated port on the ESX server for the ShoreTel VM, it doesn't like sharing a physical port with others machines.

        Comment


        • #5
          Thanks for the pointers - We will have a look tomorrow morning (UK time) and let you know how it goes...

          Our core switches are Dell PowerConnect 7048's, which we are having difficultly translating the cisco config to Dell's

          Comment


          • #6
            By team I mean in terms of the server NIC's, ie more than one NIC used for the server connection be it for redundancy/performance/both

            Also noteworthy is that you don't need to turn of spanning tree on the entire switch, it just needs to be turned off on any port connecting to ShoreTel Switches and I think also Servers (phones are OK). Note that changing spanning tree configuration during production hours could lead to service interruption as STP recalculates.

            If you are familiar with Wireshark then you could put that on your HQ server to see if you can tie any other events in with the TMS dropouts. You could also watch your switch logs for STP events.

            I used to see this (though nowhere near as severe) back in v10 but have found 12.2 to be very stable.

            It could of course not be STP - do you have a case open with ShoreTel/your partner?

            Might be worth keeping an eye on the core ShoreTel services on the HQ server as well, or take a look at the event logs if any services are restarting etc.
            Last edited by Logan1985; 04-18-2013, 10:03 PM.

            Comment


            • #7
              Well, do the switches actually lose their connection? I mean, have any users complained about anything, or have you just noticed this in Quick Look?

              Comment


              • #8
                Thanks for the response

                The TMS errors seemed to have stopped today! We didn't make any changes to Spanning tree (so may be worth reviewing anyway), but we changed our domain controller and turned off our SAN replication to a second server room via fibre connection.
                This seems to have stopped the connectivity issues for our shoretel switches. (and other issues)


                However we do want replication in the long term so need to figure out why it is causing problems - if anyone has ideas they are most welcome!

                FYI - we only have 1 NIC to the HQ.

                Users can either:
                - low not notice
                - med users are forced into release (most common)
                - high customers left in call queue while agents are idle, agents in calls are reported as release & calls can be routed to an agent already on call.

                Regards

                Comment


                • #9
                  Just to finish the thread off - we have completed various tests and the issue is caused by the fibre replication link (i.e. TMS drops when we are not even replicating). Therefore we think it is the switch set up.

                  We will keep the link down until we get a chance to correct the switch configuration, and will take into account the previous comments advise (STP off for instance).

                  If anyone has any relevant documentation that would be great, otherwise thanks for all your help

                  Regards

                  Comment


                  • #10
                    Originally posted by gr1 View Post
                    Just to finish the thread off - we have completed various tests and the issue is caused by the fibre replication link (i.e. TMS drops when we are not even replicating). Therefore we think it is the switch set up.

                    We will keep the link down until we get a chance to correct the switch configuration, and will take into account the previous comments advise (STP off for instance).

                    If anyone has any relevant documentation that would be great, otherwise thanks for all your help

                    Regards
                    Following further investigation the fibre link and switch config is fine, but our replication schedule started at the same time for the whole SAN. We have now staggered the replication procedure, and our shoreTel system is running ok plus we are successfully replicating our data store to a DR location.

                    Thanks again
                    Last edited by gr1; 05-20-2013, 07:30 AM.

                    Comment

                    Working...
                    X