I do not have an opinion about what in particular the Threats document should say (or not say) about the ORIGIN attribute. However, I am quite concerned about potential hurdles to BGPSEC deployment. It appears that breaking a current (non- standard) use of ORIGIN had little if any benefit since no one seems to use ORIGIN for its originally intended purpose. Therefore, I would like not create a new reason for some operators to deploy BGPSEC.
-Matt L. On Mar 11, 2013 6:47 PM, "heasley" <h...@shrubbery.net> wrote: > Mon, Mar 11, 2013 at 03:45:17PM -0700, Randy Bush: > > heas, > > > > > Whether anyone likes it, it has become a TE knob of sorts, in a > protocol > > > with few such knobs, and many smaller transit providers rely upon it to > > > affect route selection for non-malicious reasons, such as to balance > their > > > own transit links. Without providing an alternative, it will be > crippling > > > to many to inhibit its use as such. > > > > that is not exactly the qestion, thought i gives a hint. by "smaller > > transit providers rely on it" do you mean that the value is legitimately > > set or altered by other than the originating AS? if so, then signing it > > would be bad. if not, it could be good. so could you please clarify? > > that is exactly what I am implying/saying. > _______________________________________________ > sidr mailing list > sidr@ietf.org > https://www.ietf.org/mailman/listinfo/sidr >
_______________________________________________ sidr mailing list sidr@ietf.org https://www.ietf.org/mailman/listinfo/sidr
