Re: Experimental RFC to be: draft-stoica-diffserv-dps-01.txt

[Thomas Narten <narten@us.ibm.com>]

Hi Bob.

> I believe that diffserv already breaks the "end-to-end properties of
> IP" by rewriting DSCPs, and that diffserv changed the historic meaning
> of the TOS field in such a way as to potentially break existing uses.
> DCP promises no worse.

I don't see the two as the same at all. The IP fragmentation field,
has a well-defined use in IP. Fragmentation happens all the time.
This ID attempts to "borrow" these bits and use them for something
else (because there are no other convenient bits to use) while the
packet traverses one part of the network. Those bits must be restored
to their original meanings prior to delivery to the
destination. Failure to properly restore these bits will result in
failures (e.g., packets won't reassemble correctly).

DSCPs, on the other hand, were less clearly e2e in that their use in
practice was never that well defined or made use of. And the IETF
community (after careful thought) made an explicit decision to
redefine those bits for diffserv. I don't expect that to ever happen
with the ip_off field. Do you?

> "Serious damage" seems a bit exaggerated, but in any case, DSP offers the
> possibility of providing serious gain to the Internet and its users.  It
> might keep the telcos from imposing something you will like even less,
> like the circuits waiting in the wings!

When things go wrong, the result will be a less robust internet (e.g,
hard to diagnose failures). Is that not potentially serious damage?
(And wording you quote did include the word "potentially")

> This document seems to define an important and perhaps useful piece of
> Internet research, which might help solve a problem that the IETF has:
> QoS.  You can't have progress without at least some minor risks.

I have no opinion on the underlying QOS ideas in the document.  If
they could be discussed/tested/experimented without redefining the
ip_off field, I would be OK with that.

Thomas