odd behavior on incoming FXO line -- MGCP gateway

j.dyer · ‎03-07-2006

I have an MGCP gateway (a 2801 router, running 12.4(4)T1) used for FXO connections -- it has 3 4port FXO cards installed. 10 of the ports are configured in Callmanager 4.13. On one of the ports, whenever you make an inbound call, it rings once then gives dialtone for 8-10 seconds, then rings the phone listed in the "attendant" box in the gateway config. Outbound calls work just fine -- and the other lines all work as they're supposed to. Has anyone seen anything like this?

i checked to make sure the dial-peer for the port has a "service mgcpapp" command. I also did a debug dialpeer & it seems to check the dialpeer & find everything right at the end of that 10 second "secondary" dialtone.

also, you can't do anything with the secondary dialtone -- can't dial numbers or anything -- you just have to wait through it to get the phone ringing.

john

agopala · ‎03-07-2006

can you check if you have a route pattern that is similar or the same as the attendent number of the offending FXO port. check on the router for any such dial-peers too

j.dyer · ‎03-07-2006

the router is all mgcp, so there are just the basic dialpeers that have the "service mgcpapp" command, nothing with destination patterns, no SRST/h.323 dial peers in the box.

when i do a route plan report in callmanager, my attendant number of 6104 is the only thing that starts with 61. there are some 62 & 63 numbers, but nothing that's close to the attendant DN. No translation patterns start with 6, so nothing there. No route patterns either. I do have a pickup group of "6" but I don't think that would be a problem.

the ONLY thing different about this particular FXO from the other 9 in the router is that it's in a different partition/calling search space combo -- because only a couple phones are allowed to use it.

Erick Bergquist · ‎03-07-2006

Does the voice-port have any configuration on it like caller-id, etc? How about if you change one of the other port attendant DNs to be same DN where this port goes to see if same thing happens or not? That would help isolate problem down a bit. Also may want to try renaming that call pick up of 6 briefly and see if that helps. I have had a strange delay issue with a pilot-point or hunt pilot being delayed due to a similar number on a park number once or twice.

CCM Detailed traces will show what is going on and if there are multiple matches being considered. Have you tried DNA to see what it comes back with?

agopala · ‎03-08-2006

ok, i think we should do some isolation:

1. what happens when you put the fxo in the same partition as the rest?

2. what happens when you keep the different partition, but swap attendent DNs?

If 1 fixes the issue, I would say CCM is doing something wrong. CCM traces shold tell us the story.

If 2 fixes the issue, I would lean towards the router doing something wrong. In that case,

debug mgcp pack

debug vpm sig

debug vtsp all

should tell us what is going on.

Regards,