On Wed, 13 Mar 2019 at 14:03, William Atwood <[email protected]> wrote: > > Shyam, > > Specifically: > > Marcelo Bagnulo, Alberto Garcia-Martinez, Juan Rodriguez, and Arturo > Azcorra, "The Case for Source Address Dependent Routing in Multihoming", > International Workshop on Quality of Future Internet Services (QofIS > 2004), Lecture Notes in Computer Science (LNCS, volume 3266). > > The authors were at Universidad Carlos III de Madrid, Leganes, Madrid, > Spain at the time of writing. > > This was published in 2004, so your claim to be the first one to put > forth this idea is incorrect. >
So quite a while before 2004 there was a capability to perform "policy based routing" (PBR) available on at least one of the major routing vendors implementations. It could use the IP packet source address as a criteria to select a forwarding behaviour. For example, in 1999, I deployed PBR at a major hospital to select one of 3 site egress links, using IP source address as the selection critiera. The Linux kernel version 2.2.0, released in 1999, also gained these sorts of capabilities - the "iproute2" utilities are the interface to them. If the claimed invention is using source addresses as a forwarding input, then RFC2991, "Multipath Issues in Unicast and Multicast Next-Hop Selection", from 2000, which suggests using the packet source address as one of the inputs for link selection when ECMP, would also be an earlier example, It's an old idea. Regards, Mark. > However, as Fred says, RFC citations are based on "need to know", not > academic precedence. > > Bill Atwood > > On 12/03/2019 10:31 p.m., Fred Baker wrote: > > On Mar 12, 2019, at 1:13 AM, shyam bandyopadhyay <[email protected]> wrote: > >> I have said many times earlier that the > >> basic principle based on which RFC 8028, RFC 8043 > >> and draft-ietf-rtgwg-enterprise-pa-multihoming > >> are written, i. e. "default routing based on > >> source address of outgoing packets" was first > >> introduced on draft-shyam-site-multi. > > > > Speaking for myself, having the position of a co-author on two of the > > documents you mention. > > > > I have not inserted a reference into the documents for the simple reason > > that I was unaware of draft-shyam-site-multi and my thinking didn't depend > > on it. When I insert a reference into a paper, it's not in the academic > > sense in which one lists all possible mention of a topic and comments to > > associate with it or differentiate from it; I refer to a document to help > > explain some aspect of my thinking, or to lead a reader to some other > > technology that my thinking depends on. To give an example in the context > > of another paper, I mention that someone should validate a DNSSEC signature > > at some point, and I refer to RFC 4035 because it discusses what that means > > and how one goes about it. > > > > The earliest mention of source-dependent routing that I know of is academic > > work done by Marcelo Bagnulo at (I believe) the University of Madrid, in > > Spain. > > > > I'm sorry if that is offensive; for my part, it is pragmatic. I'm not > > writing a thesis, I'm writing a specification. I don't pretend to know > > everything. I do try to state clearly what I mean when I write something, > > and give my readers the references they need to understand and use it. > > > > > > _______________________________________________ > > rtgwg mailing list > > [email protected] > > https://www.ietf.org/mailman/listinfo/rtgwg > > > > -- > > Dr. J.W. Atwood, Eng. tel: +1 (514) 848-2424 x3046 > Distinguished Professor Emeritus fax: +1 (514) 848-2830 > Department of Computer Science > and Software Engineering > Concordia University EV 3.185 email:[email protected] > 1455 de Maisonneuve Blvd. West http://users.encs.concordia.ca/~bill > Montreal, Quebec Canada H3G 1M8 > > _______________________________________________ > rtgwg mailing list > [email protected] > https://www.ietf.org/mailman/listinfo/rtgwg _______________________________________________ rtgwg mailing list [email protected] https://www.ietf.org/mailman/listinfo/rtgwg
