Eliot, On 23 Feb 2012, at 13:47, Eliot Lear wrote:
> Damian, others, > > On 2/23/12 11:15 AM, Damien Saucez wrote: >> Hello Eliot, >> >> May I ask you why you say that LISP+ALT relies on caching? >> >> LISP+ALT is just a BGP based overlay, where is the caching >> there (is the BGP RIB that you consider as a cache?). >> > > An investigation I ran looked at the possibility of doing a "push" versus > "pull" of the mappings. It's not where the main LISP group went, but it is > possible. To be fair there is still caching of state, but the performance > profile. See draft-lear-lisp-nerd. I remember NERD very well and I think that NERD could be useful in some local environments. However, in the global Internet, NERD implies that every ITR has all the mappings and their associated churn which might impact the scalability. All that is nice but it does not answer my question, why do you say that ALT relies on caching? And what do you mean by "to be fair there is still caching of state"? Thank you, Damien Saucez > > Eliot _______________________________________________ lisp mailing list [email protected] https://www.ietf.org/mailman/listinfo/lisp
