On Thu, Mar 31, 2011 at 11:49 AM, Randy Bush <[email protected]> wrote: >> this also seems (to me) to imply that 'invalid == drop' policy is >> global, no? I suspect for a great long while 'invalid == lowered pref' >> will predominate. Hopefully when we get more comfortable and more >> reasonable with operations of the rPKI we'll have less cheese to deal >> with. > > this was my original position, but [other] operators whacked me out of > it. essentially the argument is, if you're not gonna drop something > that is clearly invalid, then why the heck run it?
I may have 2 routes, which one is better? given all things equal i'd prefer 'valid' over 'invalid' origins. I'd even prefer a longer-as-path-valid-origin over short-as-path-invalid. I don't want to shut the door completely, I like having a choice. > and the rpki having all the consistency problems of a widely distributed > database is not related to, or effected by, incremental deployment. ok _______________________________________________ sidr mailing list [email protected] https://www.ietf.org/mailman/listinfo/sidr
