Route list / Route Group with H323 VG and CUCM 7.1.3

[dtran]

Hi all,

I am running into an issue with RL / RG with H.323 VG and CUCM 7.1.3. I am looking for inputs/suggestions and hope someone can help !!!

I have 1 route list with 2 route groups in it. Each RG has 1 Voice Gateway in it. The primary VG has 4 PRI's (2 PRI's handling incoming/outgoing local calls and 2 PRI's handling incoming/outgoing LD calls). My problem is when the 2 PRI's that handle incoming/outgoing LD calls go down on the Primary Voice Gateway, LD calls fail to reroute over the backup Voice gateway which also has 2 PRI's.

Has anyone ran into this issue ? I appreciate any inputs/suggestions !!! Thanks in advance !!!

D.

[Chris Deren]

So, the call is probably attempting to go out if the local PRI on that Gw as it matches dial peers, post your config to verify. If that is the case you need to get creative with digit manipulation. If that is not the case ensure service parameter reroute on busy flag is set to true.

Chris

Sent from Cisco Technical Support iPhone App

[dtran]

Hello Chris !!! Thanks for helping me out !!! Please see the attached config !!!

[Chris Deren · Accepted solution]

So, do you have "Stop Routing on User Busy Flag" and "Stop Routing on Unallocated Number Flag" set to FALSE? You need it that way for this to work.

Looks like you do not have overlapping dial peers for LD routing, but that number are you dialing as LD?

You should change your voip dial peers with destination-pattern of .... to something more specific, somthing that at least starts with your possible incoming DID ranges, i.e. [2-5]... to avoid possible hairpining calls back to CUCM.

If you can test again please provide "debug voice dialpeer" output.

HTH,

Chris

[dtran]

Thanks Chris !!! I very much appreciate your help !!!

I'll check in CUCM make sure "Stop Routing on User Busy Flag" and "Stop Routing on Unallocated Number Flag" are set ti False. I'll run some test calls and see how that goes !!!

Yes, only LD calls were failing. Locals were fine.

D.

[Chris Deren]

I would also recommend looking at your Local dialing, it appears you local dialing is defined as certain area codes using 11 digit dialing, which overlaps with your LD pots dial peer 1, what is most likely happening is that if you send a call to GW1 and all Local circuits are busy/unavailable dial-peer 1 is matched and you are sending the call out of this GW via LD circuit which most likely is costing you more.  The only way to make it work with peer to peer protocols such as H323 and SIP is to manipulate the digits on CUCM side. For example for LD calls you prefix another 9, so when it hits the GW it would match dial peer of 991[2-9]..[2-9]...... and local calls would simple match 91[2-9]..[2-9]......

This way when particular trunk types are unavailable on one GW CUCM pulls the call back and tries another Route Group as long as you  have "Stop Routing on User Busy Flag" and "Stop Routing on Unallocated Number Flag" set to False.

HTH,

Chris

[dtran]

Hello Chris !!! I very much appreciate your inputs !!!

The local PRI's only handle tollfree and local area codes that I explicitly defined in the config and the circuits were up and available. I was able to make local and tollfree calls. Only LD calls were affected when the 2 LD PRI circuits went down for some reason LD calls fail to reroute over to the second RG in the RL. I'll have to double check the "Stop Routing on User Busy Flag" and "Stop Routing on Unallocated Number Flag"  and make sure those are set to False and test it again.

Thanks Chris !!! Very much appreciate your help !!!

D.

[Chris Deren]

As to the local call routing, I udnerstand you have explicit dial peers which means these are always going to be matched first, however you also have dial peer with desitnation pattern 9T which would match local dialing as well, so what is going to happen when you dial local number, the call is going to hist the first GW attempt to send the call out via local trunks, but if all local trunks on this GW are unavailable the call will go out via the LD circuit on this GW it will not be pulled back to CUCM for second RG routing unless Local and LD circuits on GW 1 are unavailable/busy.

Chris

[dtran]

Hello Chris !!!

I updated the parameters "Stop Routing on User Busy Flag" and "Stop Routing on Unallocated Number Flag" in CUCM to False and ran some test calls but the result is the same, calls still didn't failover to the secondary RG for some reason. When I did the test I administratively shutdown the controller T1 on the VG in the primary RG and I wonder if the VG bahave any difference when I admin down the controller T1 vs circuit goes down.

I also added this command "no dial-peer outbound status-check pots" on the VG but didn't seem to help either.

what do think could be wrong ?

Thanks Chris !!!!!

D.