[c-nsp] Rate limiting questions

[Justin Shore]

Phil Mayers wrote:
>> I'm specifically talking about upstream traffic from the client, not 
>> downstream from the HSRP routers.  Downstream will always flow out the 
>> active as expected.
> 
> Not always. The HSRP standby has an active connected route for the
> subnet and will (may) export it via a routing protocol, so the return
> path *may* in some topologies be via the standby.

That is true.  Ideally that wouldn't happen but I can see where it would 
happen even in my own network.  If our primary 7600 in the core goes 
down then traffic to our DC will be routed through the second 7600 and 
to the second DC router.  Now, in our case I am doing interface tracking 
on the interface facing the 7600s so unless the IGP melts down on the 
primary 7600 instead of the router crashing we should have an interface 
state change to work with to force HSRP to react.  Our IGP has only 
melted down 3-4 times so far this year so it's not that unlikely I 
suppose (long story).

This would be a good application for BFD-enabled HSRP.  Of course you 
first have to have BFD support in your routers, the right version of 
BFD, BFD for IS-IS and BFD for BGP (hint, hint).  Not complaining; just 
hinting strongly, maybe even begging...

Justin