@techreport{chin-nfvrg-model-dynamic-service-chain-01, number = {draft-chin-nfvrg-model-dynamic-service-chain-01}, type = {Internet-Draft}, institution = {Internet Engineering Task Force}, publisher = {Internet Engineering Task Force}, note = {Work in Progress}, url = {https://datatracker.ietf.org/doc/draft-chin-nfvrg-model-dynamic-service-chain/01/}, author = {Jonathan Chin}, title = {{A Model Based Approach to Dynamic Service Chaining}}, pagetotal = 13, year = 2016, month = feb, day = 3, abstract = {In the context of Network Function Virtualization (NFV), a service chain refers to a group of Virtual Network Functions (VNFs) deployed in a virtual infrastructure manager (VIM) environment, such as OpenStack, to apply a set of networking services to users' traffic passing through it. Depending on the service requirements, the service chain may consist of a single VNF or multiple VNFs deployed serially, or in parallel, and the Virtual Network Forwarding Graph (VNF-G) describes the topologies of how these VNFs are interconnected together via virtual links to provide different networking services. After a service chain is deployed, its user may require to modify the VNF-G from time to time, and on-demand, to meet changing service requirements. Therefore, operators may require to support a high degree of flexibility to orchestrate the provisioning and dynamic modifications of VNFs in service chains. This document presents a hierarchical model based approach to allow the NFV Orchestrator (NFV-O) and the VNF Manager (VNF-M) to support the dynamic provisioning and on-demand modifications of VNFs in service chains.}, }