RE: about QoS context

Hi all,

Just to second Rajeev's point, the important context to transfer would =
be
related to packet classification state, packet metering state, traffic=20
shaping state, etc.

Essentially what you want to do is transfer the needed contexts so that
when an IP flow is seen a the nAR, the packets are handled they same
way they were at the pAR w.r.t. QoS.

John



> -----Original Message-----
>=20
> when a packet comes up an interface, forwarding treatment (i.e., QoS) =
is
> afforded before the packet is routed. So, the packet is classified, =
metered
> (re)marked in Diffserv. There are two kinds of classifications =
recognized in
> Diffserv: multi-field and behavior aggregate. The former =
classification uses
> more than just the Diffserv Codepoint, for instance the IP addresses =
and
> port-pair.
> Also, traffic shaping at the edge is critical for Diffserv.
> On an access link, all the above operations (classification, metering =
and
> marking)
> introduce state.
>=20
> Some stuff on this at http://people.nokia.net/~rajeev/Berkeley.ppt
>=20
> -Rajeev
>=20
>=20
> [email protected] wrote:
>=20
> > Hi everybody,
> >
> > I'm sorry for my intervention. I'm currently studying the=20
> context transfer
> > approach; I find that in most of the literature, QoS is=20
> mentioned as a
> > potential candidate service. As you know there are two well=20
> known different
> > approaches: Diffserv and Intserv, but neither of them seems=20
> to match the idea
> > of the control transfer protocol. Diffserv doesn?t keep any=20
> information
> > specifically related to the MN, on the other hand, Intserv=20
> does keep state
> > information but not only in the access routers but in the=20
> entire path.
> > I?d like to know if I?m missing or misunderstanding=20
> something. (or if I?m
> > totally lost)
> > I?ll appreciate your comments.
> > Thanks a lot in advance for your time.
> >
> > Christian G.
> >
> > _______________________________________________
> > Seamoby mailing list
> > [email protected]
> > https://www1.ietf.org/mailman/listinfo/seamoby
>=20
>=20
> _______________________________________________
> Seamoby mailing list
> [email protected]
> https://www1.ietf.org/mailman/listinfo/seamoby
>=20