Recently upgraded to Unity 4.1.1. Running a dual integration with CM 4.1.3 SR3a and Avaya G3R V11.5. We migrated to the digital PIMG units with this upgrade and are still having some issues. The one that seems to be most important to solve is the intermittent choppy messages coming from the digital PIMG side. The Call Manager side has no Audio quality issues, however the PIMG messages come accross either severely broken or with long pauses in between sounds. It appears that the volume coming into the digital PIMG's (maybe from cell phones or lines with poor quality) may be causing the issue. We have adjusted the VadNoiseFloor setting in the PIMG units down to -70 seeming to make the problem a little better. We have also started to play with the dspPbxToIPGain setting in the PIMG. A setting of 6 is the loudest it will allow us to go without the DTMF tones stopping working. Are we looking in the right direction, or should we be looking elsewhere? Any help you can provide would be great.
From Unity to the PIMGs, are they are on the same subnet or switch? If you are travesing different subnets, or switches, you can have some network issues with the PIMGs.
Try forcing to the switch ports to 100F the PIMGs are plugged into.
Reboot the PIMG... never know.. it could have a memory leak....
They are all in the same subnet and an isolated Vlan. They are all set to 100 auto/full on the switch side. We could change that, but on the PIMG boxes I do not see a place to hard code to 100 full. One side would just make it worse, correct?
Unless you can isolate it to a couple ports on the PIMG. I have a had couple ports go dead on the PIMG, but it would be a little hard. I manually have to dial each port to figure out which one is getting ring no answer.. then I take that port out of the hunt group and disable it on the PIMG (take it out of service).
I always find forcing 100f instead of auto seems to work better. On the PIMG, there is no place to force it. I would probably do a manual restart on the PIMG just to clear it and see that works.
So far, it does not look as if it is a network issue, any other ideas. Our Network Engineer will not hard code 100 full on only one end. If it cannot be done on both ends, it will have to stay 100 auto/full. There are a total of 5 digital PIMG units, and the troubled calls do not seem to point at any single unit (that we can tell).
I have a cisco 4006 with 4 PIMGs connected to and it had them coded at 10f to start, then switch to 100f and they seem fine. I have not had any issues with the voice quality.
You may want to check your Unity NIC teaming to make sure that it is set to Fault Redudant and they force to 100F and not set to Auto. This will throw off the streams as well.
Your network person should also be able to monitor those ports for dropped packets if this is happening.
We are not using NIC Teaming so that is not part of the equation. Our Primary and Failover Unity servers are running 1000 Full hard coded on both ends. They also are in the dedicated Vlan that the Call Manager servers and the PIMG's are in. All also have static IP addresses. All switches/ports have QOS enabled.
I'm stumped then. Basically, the avaya is using the digital port to connect the PIMG. The PIMG makes a SIP connection to the Unity server. Unity records the message and then dumps it the messaging server. The question is better to ask is,
On the CallManager side, what if the person uses a Cisco phone to check their messages that came in on the PIMG. Does it sound the same? What could be happening is that your messeging server is having the issue and nto the PIMG, but then again.. it would be the same on both ends (ccm and pimg)
So if the WAV file has broken or garbly, then it was recorded that way. If it plays back and it's broken up, it could be either the messeging server or the PIMG or network.
My guess, you have a bad port on one of the PIMGs. This would take forever, but call each PIMG port, and leave a message to see if it breaks up.
Sorry for the late reply.
I agree testing each port is a good next step, but you also might want to have that network engineer reset counters on those switchports and just make sure you aren't pulling any errors on speed/duplex (or anything else) At least you can rule that completely out then.