Re: [tsvwg] TCP Prague's RFC 5033 guidelines status?

["Black, David" <[e-mail]>]

Posting as an individual, not as a WG chair ...

> Across threads, some of what we're discussing as "L4S" is really TCP
> Prague end-host code or algorithms, and not strictly the L4S in-network
> parts that we're currently chartered for in TSVWG.
> 
> I think 5033 is more relevant to the TCP Prague behaviors that use the
> signals coming from L4S queues, than the L4S in-network behavior by itself.

Well, L4S imposes strict requirements on end-host algorithms for traffic that uses the low-latency service.  TCP Prague is the only known transport protocol that satisfies those requirements, i.e., no other existing TCP or DCTCP code satisfies those requirements, as far as I am aware (please correct this statement if it's wrong).

For that reason, decisions about deploying L4S and TCP Prague appear to be linked, as it would be silly to deploy a network service (L4S) that cannot be used by end-systems, and end-system usage of L4S appears to require TCP Prague.

Thanks, --David

> -----Original Message-----
> From: tsvwg <[e-mail]> On Behalf Of Wesley Eddy
> Sent: Monday, November 11, 2019 9:25 AM
> To: [e-mail]
> Subject: Re: [tsvwg] TCP Prague's RFC 5033 guidelines status?
> 
> 
> [EXTERNAL EMAIL]
> 
> On 11/10/2019 3:58 PM, Holland, Jake wrote:
> > Hi l4s-arch authors,
> >
> > In a side discussion, Pete mentioned the RFC 5033 guidelines list for
> > new congestion controls, and looking through the items in Section 3, I
> > wasn't sure whether or not they were all addressed in the L4S drafts and
> > references:
> > https://tools.ietf.org/html/rfc5033#section-3
> >
> > Would it be possible to add a section that goes through the points from
> > RFC 5033 section 3 and provides pointers to the text and/or research
> > results you know of that addresses each of them?
> >
> > I think it would be very helpful for the review of the proposal to have
> > those listed out instead of trying to find them by digging through the
> > documents and references.
> >
> > I guess it would also be OK to provide such a map on the mailing list,
> > but I'll suggest that I think it would be more helpful if there's a way
> > to include it as a section in one of the drafts.
> >
> > (And if people agree this is a useful addition, should we open a new
> > issue to track it?)
> 
> 
> Across threads, some of what we're discussing as "L4S" is really TCP
> Prague end-host code or algorithms, and not strictly the L4S in-network
> parts that we're currently chartered for in TSVWG.
> 
> I think 5033 is more relevant to the TCP Prague behaviors that use the
> signals coming from L4S queues, than the L4S in-network behavior by itself.
>