[tcpm] draft-han-tsvwg-cc (was: Re: [tsvwg] Agenda requests for TSVWG@IETF101)

[Toerless Eckert <[e-mail]>]

Thanks, Michael,

On Sun, Mar 04, 2018 at 10:34:29AM +0000, Scharf, Michael (Nokia - DE/Stuttgart) wrote:
> In the IETF, I believe the expertise for this specific document would be in TCPM, which in CC. If the authors are interested in feedback on the proposed mechanism, I would recommend to ask TCPM.

TCPM might be actually a good idea, and that WG's feedback would be highly welcome.
Let me send a separate email to the chairs and ask if they could
consider presenting this in TCPM.

> From the abstract: ?????? This draft proposes a new TCP congestion control algorithm used in bandwidth guaranteed networks.  It is an extension to the current TCP standards.???

The way i would describe it: This memo defines simple extensions to TCP-Reno CC
to support and leverage a pre-known network guaranteed (minimum) bandwidth (CIR) and
pre-known maximumum needed bandwidth (PIR).  

IMHO, this work is very much orthogonal to how you would determine PIR and CIR,
the draft mentions some IMHO cool new ideas how to do this, but this CC draft
by itself would be a great benefit to application developers today (assume 
CIR/PIR API extensions to transport service -> TAPS, etc. pp).

My most simple deployment scenario is simple GUI on video devices/apps, e.g.:
in home (STB) where i could select guaranted minimum qualiy (translated
by app into CIR), and (optional) maximum quality (PIR). On the parents living
room STB CIR would be set according to 4k, and maybe in one other room to FullHD,
and the rest has to start compete for available bandwidth all the way from *yuck*
QCIF).

Of course, with more and more high-bitrate apps like AR/VR and others, the
problem to support easy policy differentiation for different apps will increase,
but i always like to look first at the most common use case that could profit today.

I feel like we have discussed this problem forever, but i am not aware of any
actual progress that app developers could use. We really need to break
through the chicken & egg circle. IMHO we first need to enable the host
stacks to do something like this, and then folks will see a lot easier the
need to have better in-network mechanisms to actually guarantee CIR.

So, i wouldn't know something more pragmatic & simple than this proposal
based on TFP reno.

> Alternatively, corresponding research could perhaps be performed in the ICCRG. ICCRG has published RFC 6077 to document some of the open research issues in this space.

Nice overview for researchers. Can't quite find from browsing a section
related to this CIR/PIR knowledge issue, admission control is mentioned
only once in passing. Let me know what you think i should specifically
look into, and we'll be happy to relate to that RFC6077 work in the draft.

> Michael
> 
> 
> From: tsvwg [mailto:[e-mail]] On Behalf Of Yingzhen Qu
> Sent: Sunday, March 04, 2018 6:55 AM
> To: [e-mail]; [e-mail]
> Cc: Thomas Nadeau <[e-mail]>
> Subject: [tsvwg] Agenda requests for TSVWG@IETF101
> 
> Dear Chairs,
> 
> A new draft (The draft was suggested by TSVWG @IETF100) was just submitted, and we???d like to request a time slot to present it @IETF101.
> 
> Title: A New Congestion Control in Bandwidth Guaranteed Network
> Presenter: Yingzhen Qu (Huawei)
> Time required (including Q/A): 10 mins
> Draft: https://datatracker.ietf.org/doc/draft-han-tsvwg-cc/
> 
> If there is any question, please kindly let us know.
> 
> Thanks,
> Yingzhen

-- 
---
[e-mail]