408 Request Timeout - Cisco Jabber

[Dean O'Meara]

Hi,

We are currently running a Proof of Concept with one of our customers and they are reporting random disconnects when in video. The customer reported last week the video disconnected at around 25 minutes and when I investigated the VCS reported "Normal Unspecified" so I suspected the Call Control has lost contact with the Jabber Client (Maybe a dropping connection at customer home) and asked them to test again.

The customer has done the test this morning and at 1 hour 9 minutes the call disconnected. I checked call history on both VCSE and VCSC and on the VCSE the call reports as "200 OK" and on the VCSC I see "408 Request Timeout" which is what is throwing me off with fault finding this?

The customer was dialling another user setup within Call Manager, the way both users logged in and placed the call to each other was over the MRA solution.

Everything else seems to work OK other than these random video call drops.

Has anyone else had this problem and aware what maybe causing this?

[Dean O'Meara]

After doing a little more digging I stumbled across the below message.

B2BUA disconnected call on the Ingress saying "Received 'Request Timeout' to mid-dialog request"

Any ideas?

[Jonathan Unger]

Could be firewall related, do you have any firewalls in the path performing SIP ALG inspection?

Proving a firewall issue can take time, and to do so requires a hefty set of logs (if you don't have access to the firewall).

In order to troubleshoot this issue fully, the following would be needed.

1. Diagnostic logging with PCAP from EXP-C
2. Diagnostic logging with PCAP from EXP-E
3. Jabber Troubleshooting log bundles from clients on both sides of the call
4. CUCM CM trace (since the endpoints are actually registered here)
5. Calling/Called numbers/URIs and start/end time of the failed call

What we would be looking for on that second call you mentioned would be a mid-call re-invite (probably a session refresh) which was not responded to in a reasonable amount of time.

A tool that can help you digest the "loggingsnapshot" files inside of the Expressway log bundles and CUCM trace files is translatorx. 

Please let us know if you would like any assistance reviewing the traces.

[ms.borikarn.th]

I have found this problem too but my network have firewall. So my firewall team already disable SIP ALG but why problem still occur?