On May 23, 2012, at 7:18 PM, Randy Bush wrote: >>>>>> Can you expound more what you mean by "aliasing" above? Do you >>>>>> mean local-as, etc. >>>>> yep >>>> That's strange. Why doesn't the following comment to the list back >>>> in March of this year count as "no comments on list to ... >>>> aliasing"? >>>> http://www.ietf.org/mail-archive/web/sidr/current/msg04093.html >>> it counted so much we spent a good bit of time on it in the april >>> meeting. we, possibly foolishly, thought we had local convergence. >> For those of us who did not attend the April interim meeting, can you >> point to where the list was asked if they agreed with the "local >> convergence"? > > i thought i just did that :) > > but, to be fair, i find the minutes lacking, to be kind. and this is > really the chairs' job. and i am nose deep in tex with a journal paper > drealine tomorrow.
I did look through the minutes and would agree that they are "lacking". Although I see some discussion of the issue under the "AS aliasing Replace AS Local AS" and "Replace as - AS migration" sections, on p. 7, all I (mainly) see are: "One way is use pcount=0" and "Going to have to be knob". It's unclear, to me, if there were (are?) other suggested solutions made and/or how the pCount=0 suggestion is thought to work, in practice. Are these written down in detail somewhere in a draft, ideally, or in the mean time can the WG chairs, or folks who suggested the idea(s), share them with the list? In particular, I'm curious about the potential security implications of now allowing anyone to turn on this "feature" and allowing them to dump anything they want into the BGPSEC_Path_Signature attribute, (assuming that any ASN they put in with pCount=0 in the BGPSEC_Path_Signature has valid signatures of course). Thanks, -shane _______________________________________________ sidr mailing list [email protected] https://www.ietf.org/mailman/listinfo/sidr
