VCS/MCU 4500 issue

Douglas Barboza · ‎09-05-2011

Hi,

I was checking the VC devices installed in my network due to calls were drop unexpectedly.

Verifing the VCS, it is reporting "500 / Routed call has been inactive too long" message.

I have that see that message when the endpoint is inactive for a long time, but this is not the case, due to users are tryng to join a conference already in use.

I thank any comments about this issue...

greetings....

Martin Koch · ‎09-07-2011

Check the involved firewall, often its a pretty fast TCP connection timeout or an application layer gw.

Also check your search rules and the signalling/call flow.

Please remember to rate helpful responses and identify

John Faltys · ‎06-05-2012

Martin,

I have the same issue for audio calls from a CUCM cluster.

I have not done too much debugging, but looking at the call history in the VCS x7.1, I see the video participants stayed in the call, but the audio participants got dropped with the error

500 / Routed call has been inactive too long

After either 30 minutes and 41 seconds or 30 minutes and 42 seconds. By looking at the logs, it seems consistent.

Call information
State	Disconnected
Start time	2012-06-01 13:20:51
Duration	30 minutes 42 seconds
Tag	48682d16-ac27-11e1-ba28-0010f31e2a10

Bandwidth
Requested	768 kbps
Allocated	4000 kbps
Route	nz_cucm-medctr01 -> nz_dgsom_vcsc

Incoming call setup parameters
Source alias 1	sip:NPANPX6556@10.86.1.11 (Url)
Target alias 1	sip:NPANPX6861@10.160.2.17:5060 (Url)
Protocol	SIP
Address	10.86.1.11:5060
Transport	TCP
Encryption type	None
Reason	Routed call has been inactive too long
Cause	500

Outgoing call setup parameters
Target alias 1	sip:NPANPX6861@10.160.2.17 (Url)
Protocol	SIP
Address	10.164.0.16:5061
Transport	TLS
Encryption type	None
Reason	Routed call has been inactive too long
Cause	500

awinter2 · ‎06-06-2012

John,

"Routed call has been inactive too long" could indicate a problem with SIP session refreshes/re-INVITES/UPDATE failing.

To troubleshoot this however, we would need to get a diagnostics log from the VCS (as well as the xconf/xstat output of the VCS) you are seeing this error on.

The diagnostics log should have the 'Network log level' set to DEBUG and you should start logging before placing the test call. Once the test call drops, you can stop logging and raise a TAC case where you provide us with the log.

Increase the SIP refresh interval timers on the VCS Configuration > Protocols > SIP > Configuration page of the VCS could help in making the calls stay up longer, but is not likely to work as a permanent fix.

- Andreas

berenyi.peter · ‎10-18-2012

Hi,

I see, that this is a quite old issue but I've just encountered possibly a variant of this error. I've got the following components: VCS 7.1, 4505 4.3(2.30), CUCM 8.6, CTS1000 1.9.1. A permanent conference is registered with a SIP URI on the VCS. Dialing out from the MCU to the CTS works properly. But when I'm trying to dial in from the CTS system to the MCU the call is dropped on the MCU with timeout, the CTS displays that the 'number is not reachable' on the screen and the VCS show the call connected. It stays connected for 30 minutes than it's dropped with '500 / Routed call has been inactive too long'. Is there any news on this issue?

Thanks

P.

John Faltys · ‎10-18-2012

Peter,

Here is how we solved the problem.

MTP Setting on - set on the Call Manager solved it, with SIP Trunk between CM and VCS

Hope it helps.

Martin Koch · ‎10-18-2012

I would be careful with MTP regards video towards the VCS, this can break more then it helps.

Any firewall in the path, especially something with ALG, L3-Inspection, ...

DNS records and ips all correct set, any strange transforms / search rules?

Config checked again towards the deployment guides?

These are things which tend to break a setup.

Please remember to rate helpful responses and identify

berenyi.peter · ‎10-19-2012

There is no firewall, and config is set according to deployment guides. I can dial into a C codec registered on the VCS without problems. I tried switching on the MTP on the trunk, but it did not help. I can dial the CTS from the MCU too.

berenyi.peter · ‎10-19-2012

Ok, this seems to be related to bug CSCty07061. Although what is strange, that if I'm calling an endpoint without the workaround applied for that bug, than the call is dropped on the VCS with 'Request terminated' while if I'm calling a conference in the MCU than the call gets stuck on the VCS.