Embedded VM is garbled intermittently on 3300

[bambamz3]

Voice Mail messages are intermittently garbled on embedded VM messages on a 3300 CXI. This site is one of eleven on a MPLS network that is having the issue. 4.2 software and Voice is a seperate VLAN than data for QoS. I am challenged in my skillset and trying hard but this is beyond me and cust has no SWA. I am including a maintenance/software log as an example.

All Maintenance/Software Logs vid1.1 10.2.2.10
0 1 2 3 4 6 8 5 7
Log Type Log Number Severity Date Time Source Description Module File Name and Line Number
Maintenance 2976 Info 2013/Mar/11 17:48:37 E2tSp IP Network Congestion Alert
Jitter from phone 192.168.34.2 > 300ms
Call duration 31 sec
JB over=0 under=68 maxJtr=305ms
number of Pkts with J > 200ms = 2 Main Maintenance;0


As I stated, this is the only site doing this and it is itermittent.

 

[bambamz3]

btt

 

[danramirez]

what about voice communications between IP sets? grabled as well?

http://www.hatkana.com

http://www.wtelecom.com

 

[Lundah]

Most likely cause is a bad embedded VM port, or possibly more than one. I've had older LX controllers do this, only fix was to replace the controller. Call each port individually and test them (dial through to an ext, leave a message, listen to a message, etc.).

 

[bambamz3]

The only voice mails that are bad are external calls. I have been using voice quality stats and there is jitter. But only on the E2t conversion do you see packet loss. Usually it is 3 to 4 packets. I can not match these packet losses to the sys/maint logs timestamps. I believe I have 2 issues--one being the network jitter and two being a hardware issue with the e2t conversion or as stated a bad port. I will go back test the port individually again to be sure.

I have had the MPLS network checked 4 times now and they claim that it is fine and Qos is right. I have had Mitel techs twice conclude it is the network. Mitel says put wireshark in and collect info which I plan to do monday with Mitels help.

With the info provided, Mitel says the logs would be producings error with DSP's or other signs if it was hardware. I disagree from past experience where they said change the harddrive and it was the cabinet. My experience is so limited , I can't argue either way. The only thing I know is that the any hardware changes will have to be done after hours and at the customers expense. He has no prob with that but as we all know, i can't guarantee a fix.

I am going back to site an hammer the VM ports.

Can anyone explain the process that a voip call follows no answer routing to embedded voice mail. Reference this to a hardware failure, not therory. From my thinking, the only fix would be replace the cabinet but again, experience field techs let me know. I just need facts to present to the customer.

 

[LoopyLou]

Since its a CXi the only E2T resources are on the main board so yes if you need to swap out the E2T bits its a cabinet swap. How many VM ports do you have because a standard CXi has something like 4 ports and if you want to expand to 16 then you need DSP resources? I suppose if you have the additional ports it could be the DSP module that is the issue.

I'd tell you a UDP joke but I'm afraid you won't get it. TCP jokes are the best because you always get them.

 

[bambamz3]

16 VM ports--I inherited the systems so I will assume by the meaning of additional DSP , that dsp was added via a card --I don't know that but it does have 16 portsset up.

 

[LoopyLou]

In your testing if you find that the trouble is on the ports 5-16 then you may need to look to the DSP.

I'd tell you a UDP joke but I'm afraid you won't get it. TCP jokes are the best because you always get them.

 

[bambamz3]

I put Wire Shark between the MPLS router and the Mitel 3300. I captured two incidents where the Wire Shark reported Jitter coming from the Main site to where I am having the trouble. The main site feeds multiple other sites and no issues have been reported from anywhere else. So I figure it has to be in the network. Network still says it not their issue. I don't believe I hardware problem now. The main site has a MXE not a CXI. Does anyone see where I am missing anything?