Re: [alto] ALTO at IETF-108: Finishing the current milestones and discussion on re-chartering the WG

Hi Jan, Vijay, Martin,

Thanks a lot for the guidance. It is extremely helpful.

Please see below.

On Sun, Jun 14, 2020 at 10:55 AM Jan Seedorf <ietf@j-f-s.de> wrote:

> Dear all,
>
> Martin (our new AD), Vijay and myself had some discussions on how to
> generally move forward with the WG and how to plan the IETF-108 ALTO
> session accordingly. We believe we should finalize the remaining
> milestones (see https://tools.ietf.org/wg/alto/charters) by or at the
> latest during the IETF-108 ALTO session. Our understanding is that
> during the weekly meetings the remaining docs are making good progress
> towards such finalization, so this goal should be realistic.
>
>
I second that the authors finish all the outstanding drafts soon.

> After IETF-108, we can either (gradually) close the WG or re-charter. In
> both cases, there should be no need for an ALTO session at IETF meetings
> to discuss progress on the currently chartered milestones. If still
> needed, such progress can happen on the mailing list.
>
> ...
>
> b) 60 minutes for presentations and discussions on potentially
> re-chartering the WG (as opposed to closing down the WG).
>
> We are soliciting presentations on re-charter proposals and thoughts.
> There will be no time allocated for presentations on drafts that are not
> currently a WG item.
>
> Needless to say, any thoughts on re-chartering are welcome on the
> mailing list at any time. We hope for a fruitful discussion at the
> online/virtual ALTO session at IETF-108 that will hopefully end up in
> some form of consensus on how to move forward with the WG.
>
>
Here is an initial list of items that I feel may help us to get discussions
started.

1. HTTP/2(3) based subscription and updates

The current incremental updates are based SSE, and now we have HTTP/2(3) to
be used. Incremental updates are important and the idea of building on top
of HTTP/2(3) was raised during IESG review.

2. Network query/flow algebra
One lesson which I learned from an early discussion is that the system can
be more flexible if it is based on a language. Recently, a framework called
Flow Algebra is emerging and can serve as a generic framework to query and
abstract generic network information.

3. Multi-resources abstraction for application-layer traffic optimization
There is an early work from SZ, Tsinghua and CMCC on their design and
initial deployment of functional networking, whose essence is to include
other resources beyond networking resources to applications. CDNi can be
considered as such an example as well. The emergence of edge computing
further motivates the need.

4. Multi-domain information aggregation, proxy, broker, and transformation
A common setting is that the traffic from a client to a server traverses
multiple networks, and hence there are many use cases. Multiple early
drafts proposed ideas to handle multi-domains. I believe that
multi-domain is also a key next step from Ingmar's award slides.

5. Cellular network information exposure
Tencent and CMCC have a draft on this topic. Sabine has a draft on this
subject as well. One highly related effort is NEF in 5G. Mobile clients are
becoming more common, and hence systematically integrating cellular network
information exposure has many good use cases.

Let me intentionally keep the list short to get the conversation started.
We have about slightly more than one month to chat more.

Richard

  - Vijay & Jan
>
>
> _______________________________________________
> alto mailing list
> alto@ietf.org
> https://www.ietf.org/mailman/listinfo/alto