[c-nsp] E1 controller - clock problems with 'line' fine with 'internal'

Ras jeekay at gmail.com
Wed Aug 22 08:13:42 EDT 2007


I've recently run into a slightly strange problem with one of my E1 circuits.

We operate a hub-and-spoke setup, where a number of lines terminate
into a single aggregation router on our side, and into a bunch of
different locations/CPEs on the remote end.

For all these lines, we have always had 'clock source line' for the E1
controller on both the aggregation router and the CPE routers. This
has worked fine and the controllers show no errors.

I've just commissioned a new line into the same aggregation router,
exactly the same equipment on the CPE side (2811, VWIC2-1MFT-G703),
exactly the same equipment on PE side (2811, VWIC2-2MFT-G703). But
this time, we were seeing continuous 'Slip Secs' (top marks to whoever
made that term up incidentally), which were also showing up as
'Errored Secs' (but crucially, never 'Errored Secs').

After much investigation and a VWIC/chassis swap later, we were in
exactly the same position.

I think tried configuring the aggregation controller (just for that
one port) with 'clock source internal' and bang all the errors
disappeared completely. It's now been running well over 48h without a
single errored second, versus 1 second per second before.

For reference, the aggregation router now has:
controller E1 0/0/1
 framing NO-CRC4
 clock source internal
 channel-group 0 timeslots 1-31

and the CPE has:
controller E1 0/1/0
 framing NO-CRC4
 channel-group 0 timeslots 1-31


Has anyone seen anything like this before and/or know what might cause
this? My telco insists that they've tested the circuit end to end and
it's working as expected (and to be fair, it is now..)

Thanks,
Ras


More information about the cisco-nsp mailing list