Re: router-router NHRP

[Dimitry Haskin <[e-mail]>]

Yakov,

> 
> Dimitry,
> 
> > > You are certainly correct. So, for example a domain that uses default
> > > for exterior routing could just carry the default route and use
> > > this default route for NHRP. NHRP doesn't alter the amount of routes 
> > > that need to be carried.
> > > 
> > > Yakov.
> > > 
> > 
> > Well.. then, IMO, the "first constraint" makes your R2R proposal
> > uninteresting.  I hope NHRP's R2R will allow to limit the number of
> > forwarding devices that need to carry a full set of routes. I believe
> > that it is possible to design R2R such that external routes (whatever
> > amount is to be carried) are carried and stored to full extend only by
> > limited number of NHRP servers that can serve routing information on
> > demand to a much larger group of a lower-end access devices.  I've thought
> > that it was the original intend of NHRP.  To boot, I don't think we need
> > to design a new routing protocol to do what you are proposing since we have
> > already protocols that can do exactly that.
> 
> We certainly failed to communicate. The R2R proposal I sent to the ROLC
> list certainly allows to accomplish what you are asking for - the
> ability of a router to augment (on demand) its existing routing
> information (acquired via conventional routing protocols) with
> additional routing information acquired via NHRP.
> 
> So, for example an OSPF router that is not in area 0 could just
> maintain routes to the destinations within its own area, and a single
> default route to all other destinations. 

What is the next hop of such a default route?

>This routing information
> could be augmented *on demand* using NHRP to discover shortcuts to
> destinations either in other areas, or even in other routing domains.
>
> Perhaps it would help if you'll give me an example of what you think
> the current NHRP R2R proposal can't do.
> 
> Yakov.
> 

OK. Let assume a *transient* routing domain with around 300 forwarding
devices (this example is not quite arbitrary but is based on an existing
FR network).  This domain must be aware of up to 40,000 external networks.
These external networks are distributed behind the forwarding devices in
such a way that no high scale aggregation is possible. 
 
Every forwarding device/router is inexpensive lower-end box that is
capable to handle a few thousand (~5000) packet streams at a time,
can support forwarding cache of a few thousand entries (10,000 max), but
unable to maintain a full set of potential target (i.e. 40,000 routes).

To make our task easier, let assume that we can convince the domain owner
to install 20-30 high end routers to handle routing load.

Can you explain how your R2R proposal will handle the described setup?
I realy try to understand your proposal and hope your explonations would help
me. Sorry if I am the only one on the list who doesn't see the obvious.

Thanks,
  Dimitry