[c-nsp] Reset an interface automatically when it goes down?

[Ed Ravin]

On Fri, Jan 26, 2007 at 11:23:19AM +0100, Wolfgang Roth wrote:
> we use a Cisco 3640 with NM-2V and VIC-2BRI-NT/T modules as VoIP/PTSN 
> gateway. The device is running a 12.4(12) enterprise image.
> 
> Sometimes unpredictable one of the BRI interfaces goes down (with ISDN 
> state 'AWAITING_ESTABLISHMENT'). Until now we weren't able to find a reason 
> for this behaviour. Resetting the interface with the 'clear interface' 
> command immediately brings the interface up again.
> 
> What we want to do now is to automatically execute a 'clear interface 
> BRIX/Y' command when this happens. Is that possible and what would be the 
> best way to achieve this (cron job, TCL script, ...)?

I'm already doing this with a cron job on a Unix host that tests for
what we think is the failure condition and then uses RANCID to log in
through a still-working router interface and run the clear int command.

But I would much rather do it in TCL since it would be self-contained on
the router.  Have not had the time to do that yet, but everyone who knows
how on this list says it can be done.  Try searching the list archives
for messages about TCL or "Event Manager", you'll find lots of useful
info and links.