[Rdma-cc-interest] The follow up actions for "An Open Congestion Control Architecture with network cooperation for RDMA fabric" related drafts

["Yuxiang (Yolanda)" <[e-mail]>]

Hi All,
          During the side meeting IETF 105, we presented that “An Open Congestion Control Architecture with network cooperation for RDMA fabric” which related to 2 drafts: Draft-zhh-tsvwg-open-architecture-00 & Draft-yueven-tsvwg-dccm-requirements-00.
         We received comments that we should divide draft-zhh-tsvwg-open-architecture-00 into 2 parts: 1. open congestion control architecture; 2. Network cooperation channel.
          So, our plan is to split draft-zhh-tsvwg-open-architecture-00<https://datatracker.ietf.org/doc/draft-zhh-tsvwg-open-architecture/> and request time to present the following at IETF-106:

1. In TSVWG, show the customer requirements (who needs this) and the benefit if we use open CC architecture to configure/negotiate different CC algorithms for different traffic flows.

2. In ICCRG, provide the test results that validate the necessity of network cooperation channel.
   Before we provide the new drafts, you could read the original drafts and let us know your thoughts on this plan. Your comments and suggestions are strongly welcomed.
           You can find the drafts here:
https://tools.ietf.org/html/draft-zhh-tsvwg-open-architecture-00
https://tools.ietf.org/html/draft-yueven-tsvwg-dccm-requirements-01

B.R.
Xiang Yu
Huawei

发件人: Paul Congdon [mailto:[e-mail]]
发送时间: 2019年7月23日 22:26
收件人: Paul Congdon <[e-mail]>
主题: New IETF non-working Group Email List for RDMA Congestion Control Interest has been created

Hello Attendees of the IETF-105 side meeting on Large Scale RDMA/HPC and other known interested parties,

As discussed in the IETF-105 side meeting, we have created the IETF non-wg email list.  Please join this to continue discussions in an open forum.  Below is the announcement and instructions.  I will be updating the email list with an appropriate welcome message and pointers to material.

A new IETF non-working group email list has been created.

List address: [e-mail]<mailto:[e-mail]>
Archive: https://mailarchive.ietf.org/arch/browse/rdma-cc-interest/
To subscribe: https://www.ietf.org/mailman/listinfo/rdma-cc-interest

Purpose:
Traditional High-Performance Computing (HPC) and Remote Data Memory Access (RDMA) networks have been relatively small scale, custom, isolated network clusters involving careful tuning and manual configuration. In-network and incast congestion is seen as a key limitations in allowing these networks to scale. This list discusses what is needed to run HPC and RDMA networks at hyperscale on cloud-style infrastructure.  Research efforts have focused on addressing gaps in congestion management, scheduling incast traffic and improving the orchestration and manageability of supporting protocols, but standardization efforts appear stalled.  This list will discuss what can be done in the IETF to support running HPC environments and RDMA protocols at hyperscale.

This list belong IETF area: TSV

For additional information, please contact the list administrators.


Regards,
Paul Congdon
[e-mail]<mailto:[e-mail]>