Re: [RAM] Incremental Deployment of LISP

[marcelo bagnulo braun <marcelo at it.uc3m.es>]

El 02/04/2007, a las 8:24, Dino Farinacci escribió:

> > I think this heavily depends on what scenario you are considering
> >
> > in LISP 1 doesn't matter, since both identifiers and locators are 
> > routable
> > So, if you do have an alternative locator for the identifier, then 
> > tunnel, else, forward directly
> >
> > similar in LISP 1.5
> >
> > In LISP 2 and 3 the issue you are considering needs to be considered 
> > imho
> >
> > probably it will induce additional latency, since you probably want 
> > to check if the dest address is an identifier.
> >
> > in LISP 3 it depends wheter you are using the push or pull model
> >
> > if you have all the identifier to locator database available, then 
> > you can check on it to verify if the destination is a identifier and 
> > which are the locators
> >
> > if you don't have it, well, then you need to make a query and find 
> > out, which will take additional delay
>
> You are right non Marcelo. I think taking the high-road by doing a 
> LISP-push model gives you the best of all worlds.

by push model, we are talking about destributing the ID to locator 
mapping database to all LISP routers?

are you thinking in using a protocol for that distribution, like BGP 
for instance?

There have been a few proposals on this area before, which were very 
interesting imho, but i guess some people though they were too complex


regards, marcelo


>  We still can incrementally deploy tunneling and by augmenting 
> LISP-CAs and LISP-peering routers we can get EID-to-RLOC integrity, 
> DOS-protection, and non-routable IDs.
>
> Mark Handley and I are working on a proposal. Give us some time to get 
> it written down.
>
> Dino


_______________________________________________
RAM mailing list
RAM at iab.org
https://www1.ietf.org/mailman/listinfo/ram