Well, from my experience it could be two likely candidates - the modem
firmware or IOS rev. I did not intentionally leave out the possibility of
hardware, but I would consider that the very last recourse - and any dummy
could suggest that!
Try this, if you have some flexibity to experiment. Keep half (or some
other reasonable proportion) of your modems on the same firmware (2.7.1.1),
and them upgrade the other half to 2.7.2.0. I have used 2.7.2.0 with good
success, even with 12 (7) T. You can do this under the SPE interface
config, if I am not mistaken. After all this, reload.
Now, take several hours to a day's worth of calls. Then, using the sh modem
output (and similar sh outputs), compare the health and success rates of the
two sets of modems. Is there a difference? Yes, then go with the better
firmware. No, then throw the router out the window (just kidding!).
If no, then consider an IOS upgrade. From my memory (please verify this, I
could be mistaken) 12 (8) T is not on the compatibility matrix from cisco
for the 2.7.2.0 firmware. What do I use? 121-4.0.1.T - yeah, that's
right!!! "ultra new at the time, almost beta code". Well, regardless of my
doubts, Cisco recommended it, and it works like a charm.
I hope this helps.
Todd
-----Original Message-----
From: Arie Vayner [mailto:ariev@netvision.net.il]
Sent: Saturday, January 13, 2001 12:53 PM
To: cisco-nsp@puck.nether.net
Subject: Problems with digital modem cards on AS5300
Hi
We have a large quantity of AS5300's for dialup user access.
Recently we have encountered a new problem which seems to have started out
of the blue:
We see the Digital Modem Cards getting stuck, so users cannot connect. When
the problem happens, we can see on 'show modem' that users cannot connect to
all the modems on a single modem card (we have 2 cards each...). We only see
modems in C mode (Connecting...), while on the other modem card there are
modems in * state. Also no modems are marked B (Bad).
We run IOS 12.0(7)T on the ASs. The modems are Mica's with 2.7.1.1 firmware
revision.
When we reload the RAS, the problem goes away for a few hours (even days)
but then comes back.
We first thought it to be a simple hardware problem, but we keep getting it
in different POPs and on too many devices...
Has anyone ever seen such a problem? Ideas?
-------------------
Arie Vayner
This archive was generated by hypermail 2b29 : Sun Aug 04 2002 - 04:12:25 EDT