Brian E Carpenter <[EMAIL PROTECTED]> writes: > > We can already classify ESP traffic. The SPI for a particular host to > > host connection is externally visible likely unique for a given > > flow in most applications. (The VPN case is a bit different but the > > VPN case is, we hope, not the bulk of traffic in the future.) > > Sorry, the SPI is no good for diffserv classification > because it has no semantics.
Neither does the flow label. Both are just a number that can be used to distinguish a bunch of traffic flowing between two hosts from other traffic flowing between two hosts. Neither has any semantics beyond that. > > We can also already label the individual packets according to > > traffic class. > > No, because it's mutable. We need sonmeting that *isn't* mutable > and *does* have semantics for classification. I am perhaps a bit thick, Brian, but all the flow label seems to indicate is that a host believes that a bunch of traffic going to another host is distinct from a bunch of other traffic by some mysterious criterion it has itself selected. Since we've imposed more or less no other implications in the draft (am I mistaken here?) it doesn't have particularly any semantics for classification there either. > > > or simply a matter of heat dissipation and the marginal cost of > > > a chip set. > > > > Well, if it truly makes no difference, why are we advocating the change? > > I didn't say it makes no difference. I said that the difference is more > likely to be in cost and waste heat than in speed. Given that the manufacturers who are doing this are already building the transistors in to walk the header chain... -- Perry E. Metzger [EMAIL PROTECTED] -- NetBSD Development, Support & CDs. http://www.wasabisystems.com/ -------------------------------------------------------------------- IETF IPng Working Group Mailing List IPng Home Page: http://playground.sun.com/ipng FTP archive: ftp://playground.sun.com/pub/ipng Direct all administrative requests to [EMAIL PROTECTED] --------------------------------------------------------------------
