On 25 Jul 2011, at 15:42, Dino Farinacci wrote: >> On Mon, Jul 25, 2011 at 3:26 PM, Dino Farinacci <[email protected]> wrote: >>> Darrel has already mentioned that the ITR can "forward-on-cache-miss" by >>> default if the implementation wants to protect its cache. >> >> How can these packets arrive at their destination? > > If destined to non-LISP sites, they go natively, just like they would in a > router today. If destined for LISP sites and sent natively because you don't > want to store the locator-set for the site (due to the cache filling up), > they go natively and will hit the closest PITR which will encapsulate to the > destination LISP site. >
If you are using ALT, you can also imagine that the PITR will "pre-fetch" the EID mappings for "all" the EID prefixes announced on the ALT. You thus have mappings for "all" the potential EID. For this kind of PITR, the only limit is the price of the box, but you can have distribution or hierarchy of such PITRs: a CDN of PITR :-D Damien Saucez > Dino > >> >> -- >> Jeff S Wheeler <[email protected]> >> Sr Network Operator / Innovative Network Concepts >> _______________________________________________ >> lisp mailing list >> [email protected] >> https://www.ietf.org/mailman/listinfo/lisp > > _______________________________________________ > lisp mailing list > [email protected] > https://www.ietf.org/mailman/listinfo/lisp _______________________________________________ lisp mailing list [email protected] https://www.ietf.org/mailman/listinfo/lisp
