Successuin to Avaya using h323 trunks

[captain99]

Hi all

I have been struggling with this problem for 3 months now and am beat.

I have in the UK a successsion with 4.0 software, sig server and h323 trunks & callpilot 4.04. Fully patched. In the USA there is an avaya (sorry don't know which type but suspect a definity) again with h323 trunks.

I can make calls to the avaya and they can call me both tdm phones (3904) and ip phones i2004 etc. very easily and with good call quality using g729 codec.

The problem is when an avaya user calls the callpilot the call is set up and all looks conected but the avaya caller hears nothing at all no speech both ways. The virtual trunk stays up for the length of time the the callers holds on or until the callpilot times out because it has not received any mf tones.

The problem happens if you call the callpilot direct or are forwarded form a phone on no relpy.

If any one have a similar set up I would be most grateful for their thoughts.


Thanks

John

 

[dlesap]

What kind of Succession you have? Generic... Give ISS from LD 22
Is the Call pilot runs fine from local Nortel Ip phone?
It looks you have a voice streaming problem or a hardware design fault.
Give details info on these points.
The reason can be found by sniffing the call. That's our new maintenance tool for everyone who wish to implement and maintain voip systems.

 

[captain99]

More info as requested

VERSION 3021
RELEASE 4
ISSUE 00 T +
IDLE_SET_DISPLAY
DepList 1: core Issue:11(06/10/16:20:00EST)

IN-SERVICE PEPS
PRS/CR PATCH REF # NAME DATE FILENAME SPECINS
Q01102996 ISS3:1OF1 p20109_3 12/12/06 p20109_3.cpt YES


LOADWARE VERSION: PSWV 111

INSTALLED LOADWARE PEPS : 0



I have 5 media cards all patched 2 sig servers again patched. The pilot also has su02 installed.


ss and media cards
RELEASE 4.00 ISSUE 55


The pilot works with any phone on site and also works from the pstn as well.

I have sniffed the tlan and the elan but cannot see the problem. However I have not yet convinced the USA end to provide a sniffer trace as the companies network dept has issues with people bringing computers onto and taking away traces from their lan.

Any thoughts however strange ?

Regards
John

 

[dlesap]

thanks for the details.
So i thing the problem is somewhere in the IP routing. Only the T-LAN have to be sniffed. In that trace you may open the UDP ports only. As you said signalisation is ok, somewhere the UDP voice packets are stopped or doesn't use the right way. You may find the voice ports number, as set up in element manager, blocked in any data network equipement. Also check for new deplist or PEP that may be avaliable. As Nortel says, in that rls your Media cards may load rls 4.00.65

 

[captain99]

If the data network was blocking the port then would the normal tdm (3904) call also have the problem?

I cannot see the difference between a callpilot call which is in essence a call to a tdm port (call pilot ports are just 2008's) and a normal 3904 call.

 

[teekow]

On the route, try turning off TRO. I have had similar problems in the past that went away when I disabled TRO on the Virtual Trunk route.

Another suggestion is to try G.711 instead of G.729 to see if it reacts differently.

Good Luck

Rob

 

[phoneguyv]

Since we're throwing things out there I'll add my .02 to the pot of suggestions.

I had an issue with calls going into CallPilot under similar circumstances. I would receive error messages every time a call was transfered out. I found it to be an issue whenever a call was put on hold. The eventual fix, change NASA to YES in the DCH of the virtual trunk.

I had to set TRO to yes to avoid other error messages as well. I saw someone mentioned changing that to NO above.

Does your trace say the call is actually established or in some kind of waiting state? Did you do a LD 80 enhanced trace to monitor the channel and see what the progress is?

V