SIP Trunk Fallback when peer not available

[CHRISTIAN END]

Hi,

i has a SIP Trunk that is pointing to a third party Server that is a callcenter. The Server sends back the calls to Cisco Phones on the CUCM where the calls come from. We have only one peer on the other side.

When the peer server on the other side will be down, routing calls to this destination will not work any longer. How can we do a fallback?

I think about a fallback that sends the calls when the peer not available  to a hunt group of Cisco Phones.

How can i manage this?

CUCM Version is 11.5.

Thanks a lot!

Christian

[Vivek Batra]

Once call has been discover its route and hits trunk/gateway, there is no provision to route the call back to the cluster.

- Vivek

[CHRISTIAN END]

There is no way? Means when the SIP Trunk is out of order, there is no fallback?

I know that i can setup a second server and point the SIP Trunk to two peers.

[Jitender Bhandari]

Hi 

When CCM routes the call towards the RL/RG/Device, and call disconnects. CCM looks into the call release code/cause value and decides whether it should further hunt any other devices in RL to route the call. If it does have the device it will try to send the setup to next idle device. The service parameters in CCM dictates whether CCM should further route the call or stop hunting.


Clusterwide Parameters (Route Plan)

Stop Routing on Out of Bandwidth Flag Required Field                   False
Stop Routing on Unallocated Number Flag Required Field            True
Stop Routing on User Busy Flag Required Field                               True
Stop Routing on Q.931 Disconnect Cause Code                             "Empty" -----You can defined the cause codes you want

CCM uses the above predefined parameters to take the decision. For eg: Call is released due to Out of BW (Cause code) , CCM looks into the parameters and will still hunt the other device to route the call, as it is set to "False".

So if the other side responds with a valid disconnect cause CUCm would be able to chose next Trunk.

Also you can enable option ping to query the far end is UP\DOWN, with 3rd party server i am not sure about the compatibility but it is worth a shot.

(Rate if it helps)

JB

[Vivek Batra]

What I meant is there is no way to route the call back to the cluster in such cases.

You always have an option to route the call to other peer outside your cluster when primary peer goes down.

- Vivek

[CHRISTIAN END]

I got a dirty solution. But i want to discuss this with any of you here. 

From this discussion, i setup the solution for me and it works very well. But can it have some issues, that i did not think about? Loop or something else? The incoming CSS for the SIP Trunk to CUCM itself will have only access to the Hunt Pilot and nothing more.

https://supportforums.cisco.com/discussion/12310986/cucm-91-and-third-party-call-center-integration-over-sip-trunk