[Irtf-rr] AW: AW: Differentiated Routing, not only plain rambo-SPF

[Hummel Heinrich]

I am sorry, but there is no signalling (like RSVP, RSVP-TE, CR-LDP) for =
doing DiffRout.

Based on the information exchanged by the routing protocol (like OSPF) =
all nodes should be able to
compute the same smallest size tree which interconnects the edge =
routers (those with attached users or attached external interfaces), =
and which complies to some specific constraints =
(QoS/BW/Policy,eventually avoiding overlaps with other=20
road systems),and which are associated to the same DSCP(s).  // plural: =
see my  previous email from today

But there is no signalling as to "establish the DSCP-associated road =
system" !!!

However,it would be a fancy but completely new task to determine and =
establish an MPLS-road system,
which may lock MPLS-traffic into such slim Smallest Size Tree graphs. =
See all my work on establishing
the "road systems for VPNs" based on elementary+ hierarchical LSPs.

Heinrich



-----Urspr=FCngliche Nachricht-----
Von: Ayyasamy, Senthilkumar (UMKC-Student) [mailto:saq66@umkc.edu]
Gesendet: Dienstag, 13. August 2002 16:40
An: Naidu, Venkata; Hummel Heinrich; Manfredi, Albert E
Cc: routing-discussion@ietf.org; irtf-rr@puck.nether.net
Betreff: RE: AW: Differentiated Routing, not only plain rambo-SPF



> -> yes,it is similar;
> -> but here it is about state-less IP forwarding, i.e. about=20
> -> forwarding the individual IP packets, without any
> -> RSVP path and without any MPLS path.
>=20
>   Oh! Ok...you would like to do Crankback in
>   connection less IP environments?!?
>=20
>   In such a case, you might find below paper interesting:
> http://www.ecse.rpi.edu/Homepages/shivkuma/research/papers/te-
> latest.pdf
 I have studied this paper. Your are mis-interpreting here. The RPI
paper is an increment to OSPF or BGP for some traffic engineering
functionalties. It doesnt require signaling whereas, diffrout=20
requires signaling. Also, we are taking more from a diffserv=20
point of view.
 As i told, routing in peer to peer networks and some overlay=20
systems can use diffrout concept. We can explore certain non-
cooperative algorithms for an alternative to SPF. But as mentioned
before, this will not work out if we see from an engineering=20
perspective.=20

>=20
>   There was some discussion in E2E list regarding
>   pros & cons of this approach. I couldn't find the URL
>   but I vaguely remember that the subject of the discussion
>   is "EC++N" or something like that...