[c-nsp] Logging event link-status issues

[John Neiberger]

On Thu, Aug 26, 2010 at 4:22 PM, John Neiberger <jneiberger at gmail.com> wrote:
> We've tried to enable this on individual interfaces, but we still
> always seem to get link up/down messages for other interfaces. I see
> that there is a global "no logging event link-status" command, which
> strangely has an extra optional keyword "default". I don't understand
> why you need the "default" keyword for a global command. Wouldn't the
> fact that it is global already make it default?
>
> Anyway, my question is this: if we add the global command "no logging
> event link-status default" and then add "logging event link-status" on
> the important interfaces, then add "snmp-server enable trap linkdown",
> will we only get link down traps for the important interfaces?
>
> I know other people have run into weirdness related to this, so I
> thought I would ask here.
>
> Thanks,
> John
>

To make matters more interesting (annoying), last night we got an snmp
trap from a 7600 saying a critical interface had gone down. I and
another engineer logged into the router and saw absolutely no
problems. There wasn't even anything in the logs from the past three
hours. No interfaces had gone down at all. So what the heck triggered
the trap?

The only potential lead I've found is a bug related to EEM, but it
doesn't exactly fit our circumstances. Still, I wonder if something
happened with EEM and it sent out an errant trap.