Re: [tsvwg] Deprecating RFC 3168 for future ECN experimentation

daihuichen <daihuichen@huawei.com> Wed, 31 March 2021 12:21 UTC

Return-Path: <daihuichen@huawei.com>
X-Original-To: tsvwg@ietfa.amsl.com
Delivered-To: tsvwg@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 9E1ED3A268E for <tsvwg@ietfa.amsl.com>; Wed, 31 Mar 2021 05:21:13 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.919
X-Spam-Level:
X-Spam-Status: No, score=-1.919 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_MSPIKE_H3=-0.01, RCVD_IN_MSPIKE_WL=-0.01, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 9fppch8jffuK for <tsvwg@ietfa.amsl.com>; Wed, 31 Mar 2021 05:21:08 -0700 (PDT)
Received: from frasgout.his.huawei.com (frasgout.his.huawei.com [185.176.79.56]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 863A83A2687 for <tsvwg@ietf.org>; Wed, 31 Mar 2021 05:21:05 -0700 (PDT)
Received: from fraeml709-chm.china.huawei.com (unknown [172.18.147.206]) by frasgout.his.huawei.com (SkyGuard) with ESMTP id 4F9QJk30G1z685QX; Wed, 31 Mar 2021 20:16:06 +0800 (CST)
Received: from dggeme701-chm.china.huawei.com (10.1.199.97) by fraeml709-chm.china.huawei.com (10.206.15.37) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA256) id 15.1.2106.2; Wed, 31 Mar 2021 14:21:01 +0200
Received: from dggeme754-chm.china.huawei.com (10.3.19.100) by dggeme701-chm.china.huawei.com (10.1.199.97) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA256_P256) id 15.1.2106.2; Wed, 31 Mar 2021 20:20:59 +0800
Received: from dggeme754-chm.china.huawei.com ([10.6.80.77]) by dggeme754-chm.china.huawei.com ([10.6.80.77]) with mapi id 15.01.2106.013; Wed, 31 Mar 2021 20:20:59 +0800
From: daihuichen <daihuichen@huawei.com>
To: "Rodney W. Grimes" <ietf@gndrsh.dnsmgr.net>
CC: Jonathan Morton <chromatix99@gmail.com>, "tsvwg@ietf.org" <tsvwg@ietf.org>
Thread-Topic: [tsvwg] Deprecating RFC 3168 for future ECN experimentation
Thread-Index: AQHXImHNpOplzDSttk6NxQ3/h/GzJ6qdcOjQ//99agCAAIbpwIAADEyAgACGqXA=
Date: Wed, 31 Mar 2021 12:20:59 +0000
Message-ID: <f86230793357484fa2a3f758e3be3ad9@huawei.com>
References: <28cbb29662b344a8a67ff7a87e74b1cf@huawei.com> <202103311205.12VC5tjq077006@gndrsh.dnsmgr.net>
In-Reply-To: <202103311205.12VC5tjq077006@gndrsh.dnsmgr.net>
Accept-Language: en-US
Content-Language: zh-CN
X-MS-Has-Attach:
X-MS-TNEF-Correlator:
x-originating-ip: [10.108.242.228]
Content-Type: text/plain; charset="gb2312"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
X-CFilter-Loop: Reflected
Archived-At: <https://mailarchive.ietf.org/arch/msg/tsvwg/maHrFEHLzjzrytZf8zz5bA5Tdd0>
Subject: Re: [tsvwg] Deprecating RFC 3168 for future ECN experimentation
X-BeenThere: tsvwg@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Transport Area Working Group <tsvwg.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/tsvwg>, <mailto:tsvwg-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/tsvwg/>
List-Post: <mailto:tsvwg@ietf.org>
List-Help: <mailto:tsvwg-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/tsvwg>, <mailto:tsvwg-request@ietf.org?subject=subscribe>
X-List-Received-Date: Wed, 31 Mar 2021 12:21:14 -0000

Rod,

Mostly RED/ECN queues with very small marking thresholds, e.g., Kmin=5KB and Kmax=120KB, the exact setting differs from algorithm to algorithm. 
The purpose is to achieve very high link utilization (close to 100%) and at the same time maintains small queue size to minimize queueing delay.
The RoCE clusters in cloud have been using RED/ECN for years, and at scale.


A new algorithm even propose to use WRED to speed up the first RTT transmission rate.
https://dl.acm.org/doi/10.1145/3387514.3405878


Regards,
/Huichen

-----Original Message-----
From: Rodney W. Grimes [mailto:ietf@gndrsh.dnsmgr.net] 
Sent: 2021年3月31日 8:06 PM
To: daihuichen <daihuichen@huawei.com>
Cc: Jonathan Morton <chromatix99@gmail.com>om>; tsvwg@ietf.org
Subject: Re: [tsvwg] Deprecating RFC 3168 for future ECN experimentation

Huichen,

What type of queues are used for the low-latency ROCE traffic?

> Jonathan,
> 
> Only the queue for low-latency traffic, e.g., roce packets, is ECN enabled.
> The rest queues are usually FIFOs with tail-drop policy. 
> 
> /Huichen
> 
> -----Original Message-----
> From: Jonathan Morton [mailto:chromatix99@gmail.com] 
> Sent: 2021?3?31? 11:19 AM
> To: daihuichen <daihuichen@huawei.com>
> Cc: Steven Blake <slblake@petri-meat.com>om>; tsvwg@ietf.org
> Subject: Re: [tsvwg] Deprecating RFC 3168 for future ECN experimentation
> 
> > On 31 Mar, 2021, at 6:15 am, daihuichen <daihuichen@huawei.com> wrote:
> > 
> > Today, ECN has been widely deployed in data center networks?
> 
> May I ask, purely out of idle curiosity, whether those deployments are primarily of the single-queue variety, or of the flow-isolating type?
> 
>  - Jonathan Morton
-- 
Rod Grimes                                                 rgrimes@freebsd.org