Re: [Detnet] suggested text revision to draft-ietf-detnet-scaling-requirements
Peng Liu <liupengyjy@chinamobile.com> Fri, 30 June 2023 02:48 UTC
Return-Path: <liupengyjy@chinamobile.com>
X-Original-To: detnet@ietfa.amsl.com
Delivered-To: detnet@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 6F7A1C14CE30 for <detnet@ietfa.amsl.com>; Thu, 29 Jun 2023 19:48:19 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.888
X-Spam-Level:
X-Spam-Status: No, score=-6.888 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_HI=-5, RCVD_IN_ZEN_BLOCKED_OPENDNS=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_KAM_HTML_FONT_INVALID=0.01] autolearn=ham autolearn_force=no
Received: from mail.ietf.org ([50.223.129.194]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id XHYckmHCOTKh for <detnet@ietfa.amsl.com>; Thu, 29 Jun 2023 19:48:18 -0700 (PDT)
Received: from cmccmta2.chinamobile.com (cmccmta2.chinamobile.com [221.176.66.80]) by ietfa.amsl.com (Postfix) with ESMTP id AA516C14CF05 for <detnet@ietf.org>; Thu, 29 Jun 2023 19:48:14 -0700 (PDT)
X-RM-TagInfo: emlType=0
X-RM-SPAM-FLAG: 00000000
Received: from spf.mail.chinamobile.com (unknown[172.16.121.87]) by rmmx-syy-dmz-app05-12005 (RichMail) with SMTP id 2ee5649e425d851-27f50; Fri, 30 Jun 2023 10:47:59 +0800 (CST)
X-RM-TRANSID: 2ee5649e425d851-27f50
X-RM-TagInfo: emlType=0
X-RM-SPAM-FLAG: 00000000
Received: from CMCC-LP (unknown[10.2.53.125]) by rmsmtp-syy-appsvrnew04-12029 (RichMail) with SMTP id 2efd649e425b37f-8c42e; Fri, 30 Jun 2023 10:47:58 +0800 (CST)
X-RM-TRANSID: 2efd649e425b37f-8c42e
Date: Fri, 30 Jun 2023 10:48:42 +0800
From: Peng Liu <liupengyjy@chinamobile.com>
To: liyizhou <liyizhou@huawei.com>
Cc: "david.black" <david.black@dell.com>, detnet <detnet@ietf.org>, ryoo <ryoo@etri.re.kr>
References: <1a42fb222b8c47a58094c6f990aab950@huawei.com>
X-Priority: 3
X-GUID: 135641A1-08C8-4B10-AACF-D49FF44368E0
X-Has-Attach: no
X-Mailer: Foxmail 7.2.21.453[cn]
Mime-Version: 1.0
Message-ID: <2023063010484240842142@chinamobile.com>
Content-Type: multipart/alternative; boundary="----=_001_NextPart631177124733_=----"
Archived-At: <https://mailarchive.ietf.org/arch/msg/detnet/FQsq0GYyj_-BSTPmjhHpvwjyn9M>
Subject: Re: [Detnet] suggested text revision to draft-ietf-detnet-scaling-requirements
X-BeenThere: detnet@ietf.org
X-Mailman-Version: 2.1.39
Precedence: list
List-Id: Discussions on Deterministic Networking BoF and Proposed WG <detnet.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/detnet>, <mailto:detnet-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/detnet/>
List-Post: <mailto:detnet@ietf.org>
List-Help: <mailto:detnet-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/detnet>, <mailto:detnet-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 30 Jun 2023 02:48:19 -0000
Hi Yizhou, Thank you. I have been working on the issues of this section to address the comments of traffic class. Considering your text, I propose to change Section 3.4 as follows(most of them are in the second paragraph): OLD: The deterministic network may have the large number of traffic flows. According to [RFC2475], per-flow state identification in the network becomes infeasible as flow count scales up. So, it is almost impossible to identify individual IP flows at the DetNet data plane for a massive number of flows. DetNet allows the leverage of the flow aggregation, while individual flows may join and exit the aggregated flow rapidly in the scaling network with large number of flows, which causes the dynamic in identification of the aggregated DetNet flow. The wildcards and value ranges used in the identification may have to change in order to ensure the aggregated flows have compatible deterministic characteristics. Moreover, for scaling network, proper provision at the control plane to accommodate such higher aggregation is required. The deterministic latency forwarding mechanisms MUST scale to networks of significant size with the massive traffic flows. Meanwhile, the traffic that requires deterministic networking can significantly fill up the capacity of a link or the portion of the link which is dedicated to such traffic, for example, more than 75% and/or up to near 100% utilization. The over-provisioning of link capacity does not work in such cases. In order to guarantee deterministic latency and jitter in this environment, it is required to provide scalable queuing solutions to improve the bandwidth utilization. For instance, when the bandwithd utilization is high, the guard band in each cycle in [IEEE802.1Qch] is a type of over- provisioning and can be improved with more scalable queuing add-ons. NEW: The deterministic network may have the large number of traffic flows. According to [RFC2475], per-flow state identification in the network becomes infeasible as flow count scales up. So, it is almost impossible to identify individual IP flows at the DetNet data plane for a massive number of flows, neither for the per-node state machine configuration. DetNet allows the leverage of the flow aggregation, while individual flows may join and exit the aggregated flow rapidly in the scaling network with large number of flows, which causes the dynamic in identification of the aggregated DetNet flow. The wildcards and value ranges used in the identification may have to change in order to ensure the aggregated flows have compatible deterministic characteristics. For scaling network, proper provision at the control plane to accommodate such higher aggregation is required. Provisioning on the aggregated flows normally improve the scalability at the cost of coarse granularity of the incoming traffic filtering and protection. It is desirable that adding a flow to the network does not affect the latency of the existing flows, or requires the complex re- calculation, it should require as less work as possible. For instance, only the filtering and policing configuration on the ingress node but not the intermediate nodes. The [IEEE802.1Qbv] uses traffic class to divide the flows and the number of it is usually 8, so that the forwarding mechanisms itself isn't complex with a large number of flows or higher aggregation. However, when adding new flows, the Gate Control List may be changed, and the re-calculation is complex. There might be the method to simplify the calculation or configuration, which need more work to enhance it. Meanwhile, the traffic that requires deterministic networking can significantly fill up the capacity of a link or the portion of the link which is dedicated to such traffic, for example, more than 75% and/or up to near 100% utilization. The over-provisioning of link capacity does not work in such cases. In order to guarantee deterministic latency and jitter in this environment, it is required to provide scalable queuing solutions to improve the bandwidth utilization. For instance, when the bandwithd utilization is high, the guard band in each cycle in [IEEE802.1Qch] is a type of over- provisioning and can be improved with more scalable queuing add-ons. Regards, Peng liupengyjy@chinamobile.com From: Yizhou Li Date: 2023-06-29 20:59 To: Peng Liu CC: david.black; detnet; ryoo Subject: [Detnet] suggested text revision to draft-ietf-detnet-scaling-requirements Hi Peng, Based on some discussions in the interim, I’d suggest to add the following text to section 3.4 of draft-ietf-detnet-scaling-requirements in some place you think appropriate. It is generally desired that no per-flow, per-node state machine configuration is required for better scalability. Provisioning a new flow should require as less work as possible, e.g. only the filtering and policing configuration on the ingress node but not the intermediate nodes. Provisioning on the aggregated flows normally improve the scalability at the cost of coarse granularity of the incoming traffic filtering and protection. In addition, it is desirable that adding a flow to the network does not affect the latency of the existing flows, or that the bounded latency requirement of the existing flows still meets without the complex re-calculation after adding a new flow. Thanks, Yizhou