Tutorial

1. CFN-Dyncast Overview [EN]

Publication URL: PDF - CFN-Dyncast Overview

Introduction:

Compute First Networking (CFN) leverages both computing and networking status to help determine the optimal edge among multiple edge sites with different geographic locations to serve a specific edge computing request.

2. Innovation and Practice of CFN Based on IPv6+ [CN]

Publication URL: PDF - Innovation and Practice of CFN Based on IPv6+

Introduction:

This slide introduces the innovation and pratice of CFN based on IPv6+ technologies.

3. Innovation of Computing-Aware Routing in Intelligent IP Network

Publication URL: PDF - Innovation of Computing-Aware Routing in Intelligent IP Network

Introduction:

This slide introduces the innovation of computing-aware routing.

Paper

1. IEEE WCNC 2021

Publication URL: Paper - CFN-dyncast: Load Balancing the Edges via the Network

Publication URL: Slides - Jianwei Mao (speaker)

Publication URL: Meeting Video - Bing Liu (Remy) (speaker)

Introduction:

Multi-access Edge Computing (MEC) is a promising business paradigm in the 5G network. The principle “the nearest is the best” may not apply in some cases. How to break the boundaries among individual MEC sites and leverage the computing resources as an integrity, is the key to improve the user experience and improve the usage efficiency of computing and network resources. This paper proposes CFN-dyncast, a distributed technique that dispatches clients’ demands to an optimal site according to the load of each computing site and the network status. This paper also introduces the related design considerations, the implementation and the evaluation comparing to other load balancing techniques.

IETF

Drafts

1. Dynamic-Anycast (Dyncast) Use Cases & Problem Statement

Publication URL: https://datatracker.ietf.org/doc/html/draft-liu-dyncast-ps-usecases

Introduction:

The document identifies several key areas which require more investigations in terms of architecture and protocol to achieve balanced computing and networking resource utilization among edges providing the services.

2. Dynamic-Anycast (Dyncast) Requirements

Publication URL: https://datatracker.ietf.org/doc/html/draft-liu-dyncast-reqs

Introduction:

This draft provides requirements for an architecture addressing the problems outlined in the use case and problem statement draft for Dyncast.

3. Dynamic-Anycast Architecture

Publication URL: https://datatracker.ietf.org/doc/html/draft-li-dyncast-architecture

Introduction:

This document describes a proposal for an architecture for the Dynamic-Anycast (Dyncast). It includes an architecture overview, main components that shall exist, and the workflow.

An example of workflow is provided, focusing on the load-balance multi-edge based service use-case, where load is distributed in terms of both computing and networking resources through the dynamic anycast architecture.

4. BGP NLRI App Meta Data for 5G Edge Computing Service

Publication URL: https://datatracker.ietf.org/doc/html/draft-dunbar-idr-5g-edge-compute-app-meta-data

Introduction:

This draft describes a new BGP Network Layer Reachability Information (BGP NLRI) Path Attribute, AppMetaData, for egress router to advertise the running status and environment of the directly attached 5G Edge Computing servers. The AppMetaData can be used by the ingress routers in the 5G Local Data Network to make intelligent path selection for flows from UEs. The goal is to improve latency and performance for 5G Edge Computing services.

5. OSPF extension for 5G Edge Computing Service

Publication URL: https://datatracker.ietf.org/doc/html/draft-dunbar-lsr-5g-edge-compute-ospf-ext

Introduction:

This draft describes an OSPF extension for routers to advertise the running status and environment of the directly attached 5G Edge Computing servers. The AppMetaData can be used by the routers in the 5G Local Data Network to make intelligent decisions to optimize the forwarding of flows from UEs. The goal is to improve latency and performance for 5G Edge Computing services.

6. IP Layer Metrics for 5G Edge Computing Service

Publication URL: https://datatracker.ietf.org/doc/html/draft-dunbar-ippm-5g-edge-compute-ip-layer-metrics

Introduction:

This draft describes the IP Layer metrics and methods to measure the Edge Computing Servers running status and environment for IP networks to select the optimal Edge Computing server location in 5G Edge Computing (EC) environment. Those measurements are for IP network to dynamically optimize the forwarding of 5G edge computing service without any knowledge above IP layer.

7. IPv6 Solution for 5G Edge Computing Sticky Service

Publication URL: https://datatracker.ietf.org/doc/html/draft-dunbar-6man-5g-edge-compute-sticky-service

Introduction:

This draft describes the IPv6-based solutions that can stick an application flow originated from a mobile device to the same ANYCAST server location when the mobile device moves from one 5G cell site to another.

8. Providing Instance Affinity in Dyncast

Publication URL: https://datatracker.ietf.org/doc/html/draft-bormann-dyncast-affinity

Introduction:

The present short document defines a way to provide instance affinity that does not require, but also does not rule out per-application state.

9. Use Cases for Computing-aware Software-Defined Wide Area Network(SD-WAN)

Publication URL: https://datatracker.ietf.org/doc/html/draft-zhang-dyncast-computing-aware-sdwan-usecase

Introduction:

SD-WAN is aware of the computing power of applications deployed in the multiple sites of enterprise and can perform the routing policy according to such information. This is defined as the computing-aware SD-WAN.This document describes the use cases for computing-aware Software-Defined Wide Area Network(SD-WAN).

10. Use Cases of Computing-aware Service Function Chaining (SFC)

Publication URL: https://datatracker.ietf.org/doc/html/draft-zhang-computing-aware-sfc-usecase

Introduction:

Multiple occurrences of the same service function(SF) can exist in the same administrative domain and each occurrence of SF is called SF instance. A Service Function Path(SFP) is determined by composing selected SF instances and overlay links. The SF instances are selected according to the computing power of SFs in addition to the network information and this is defined as the computing-aware SFC in this document. This document describes the use cases for computing-aware Service Function Chaining(SFC).

11. Computing Resource Modeling for CAN

Publication URL: https://datatracker.ietf.org/doc/html/draft-liu-can-computing-resource-modeling

Introduction:

This document describes the considerations and potential architecture of modeling the computing resource in the Computing-Aware Network(CAN).

Moreover, the network and application based modeling are also presented in this document to meet the potential requirements of integrated and hierarchical modeling.

Meetings

1. CFN-dyncast Side Meeting @IETF109

Publication URL: https://github.com/dyncast/ietf109

2. CFN-dyncast Side Meeting @IETF110

Publication URL: https://github.com/dyncast/ietf110

3. CAN BoF @IETF 113(Non WG forming)

Agenda URL: https://datatracker.ietf.org/doc/agenda-113-can/

Material URL: https://datatracker.ietf.org/meeting/113/session/can

4. CAN BoF @IETF 115(WG forming)

Agenda URL: https://datatracker.ietf.org/doc/agenda-115-can/

Material URL: https://datatracker.ietf.org/meeting/115/session/can

CCSA

1. Research on Computing Measurement and Computing Modeling based on Computing and Network convergence

Introduction:

The development of computing power and network convergence is the trend of edge computing and network servitization in the future. As the basis of computing-network convergence, this topic studies how to measure computing power, how to establish a unified computing power function model and computing power demand model, and establish an evaluation system of computing power measurement. It includes the modeling method and description language of computational power, which can be measured, evaluable and mapped.

The calculation force specifically includes: (1) measurement of computing power of heterogeneous hardware; (2) measurement of computing power requirements of diversified algorithms; (3) measurement of computing power requirements of users.

The computing power modeling mainly includes: (1) Compute resource model description (2) Compute requirement model description (3) Compute power evaluation standard and test standard model.

2. Requirements and architecture of computing first network

Introduction:

Focus on carriers’ MEC bearer networks in the early stage, construct MEC computing networks in fixed network and mobile network scenarios using CFN, and research and develop a series of standards, such as CFN requirements, functional features, technical architecture, and protocols. In the future, the computing network will be extended to other scenarios, such as data centers.

3. Architecture and technical requirements of computing first network

Introduction:

This project will specify the architecture and technical requirements of the computing first network (CFN).

4. Technical requirements of routing protocol for computing first network

Introduction:

This project will specify the technical requirements of routing protocols for computing first network.

5. Technical requirements of network controller for computing first network

Introduction:

This project specifies the technical requirements of network controller for computing first network.

6. Technical requirements of identifier resolution for computing first network

Introduction:

This project specifies the technical requirements of identifier resolution for computing first network.

7. Research of openness capability of computing first network

Introduction:

This project will research the aspects of openness capability of computing first network.

8. Computing Network technologies research of micro-service scenario based upon SRv6 & SFC

Introduction:

This project will focus on the following topics:

9. Key technologies of computing-aware networking towards computing resources of whole network

Introduction:

Compute-aware networks (CAN) can achieve efficient coordination and optimal scheduling of network-wide computing resources. This research project analyzes the requirements of the Compute-aware Network (CAN), combs the key characteristics of the Compute-aware Network, establishes a network-oriented computing-awareness and scheduling system, and proposes a control plane. The forwarding plane and management plane collaborate with each other to achieve convergence of computing power and routing. Specifically, research the unified computing power management and scheduling on the control plane, and implement the computing-oriented routing advertisement mechanism; Research the sensing and measurement mechanism of computing power in forwarding plane. This paper studies the unified planning and orchestration of heterogeneous computing resources to form the functional architecture of computing-aware networks.

10. Architecture and technical requirements of computing-aware networking

Introduction:

This project will develop and research the architecture and functional technical requirements of the CAN, including the functional architecture of the CAN, functional description of the CAN, and technical requirements of the CAN.

The details are as follows:

  1. Overview of the computing power-aware network
  2. Application scenarios of the computing power-aware network
  3. Functional architecture and function description of the computing power-aware network
  4. Technical requirements of the computing power-aware network

Demo

OpenSource

News

1. CFN enables Smart Security Protection solution in the network of China Unicom(Beijing)

In the 5th Future Network Development Conference, China Unicom Research Institute, China Unicom(Beijing) and Huawei published this practice in the network of China Unicom(Beijing).

CFN现网实践发布,华为,联通研究院,北京联通

IETF Mailing List

CFN Dyncast

Address: dyncast@ietf.org

Archives: https://mailarchive.ietf.org/arch/browse/dyncast/

To subscribe: https://www.ietf.org/mailman/listinfo/dyncast

CAN

Address: can@ietf.org

Archives: https://mailarchive.ietf.org/arch/browse/can/

To subscribe: https://www.ietf.org/mailman/listinfo/can