[cisco-voip] possible VWIC-1MFT-T1 failure

[Mike Armstrong]

We had a similar problem a year and a half ago.  Check the error rates via 
the "sho service-module <interface> perf" command on the 3640.  If you're 
seeing high error rates (we saw thousands per 15-minute interval, detected 
only by the VWIC (Telco HTU-R and T-Berds reported zero errors), then you 
may have our problem, for which there's a cheap but inelegant solution. 
Here's a slightly edited version of what I reported (to this list, I think), 
back then; if you want more gory details, e-mail me off-list:

Cisco case number was E353757 -- I think any CCIE should be able to access 
it, but it won't help much.

Symptoms were noise on the inbound side of the conversation and high error 
rates -- the MFT reported line errors (we saw hundreds of thousands in a 
15-minute interval) not detected by Telco monitoring or test equipment. 
(Or by an AS5300, which we used successfully for a couple of months while 
Cisco tried to find the problem.)   We replaced VWICs, routers, IOS, 
microcode, etc., and HTUs to no avail.  Cisco eventually decided this was a 
VWIC-1MFT-T1 problem (DDTS CSCec43050).  This apparently occurs very rarely, 
and is  felt to be a Telco line-noise problem (I agree), but since the 
errors are not detected by typical Telco test equipment (e.g. T-Berd), they 
won't be inclined to diagnose or fix it.  The solution was to insert an OEM 
CSU between the output of the HTU-R "SmartJack" and the MFT to act as a line 
conditioner.  Since installing one of these, we have seen zero errors.

I've forgotten the brand of the CSU Cisco sent us -- I hadn't heard of 
them -- but I suspect any would do the job.  I was ready to buy an Adtran, 
since the Adtran folks had been fairly cooperative when we were trying to 
find the problem (Verizon uses Adtran's "Smartjack").

Mike Armstrong
UF/IFAS CREC
Lake Alfred, FL