The SDN/NFV puzzle and its missing pieces_ Pedro A. Aranda Gutiérrez Telefonica I+D October 2015
Introduction Standardisation vs. Delivery
An uneven battle with uneven arms The NSP Cycle The CAP Cycle Idea!! AVAILABLE Idea!! AVAILABLE Network Service Providers Demand Deploy Content and Application Providers Develop Deploy Publish 2-6 Months Equipment Vendors Drive Sell SDOs Critical mass of supporters Standardise Implement 2-6 Years 2-6 years 2-6 months
Well, it s not black and white Idea!! Network Service Providers DEPLOYMENT 1 DEPLOYMENT 2 Drive DEPLOYMENT 3 Vendors IMPLEMEATION 1 IMPLEMENTATION 2 IMPLEMATATION 3 IMPLEMEATION 1 IMPLEMENTATION 2 IMPLEMATATION 3 Implement SDOs STANDARDISATION Critical mass of Supporters CO NSE NSU S FINAL VERSION Standardise
Overcoming the gap SDN and NFV for rapid network evolution
The Three Gears Leads to agility Reduces CAPEX and OPEX Network Functions Virtualisation Software Defined Networking Creates abstractions to enable faster innovation Open Innovation Creates competitive supply of innovative applications by third parties 6
The Road to a modern approach to SDN We are applying concepts and tools related to software development to SDN Supporting a complete design flow High-level definition and manipulation Validation from simulation to actual debugging Beta versions by slicing Phased deployment Integrate virtualized and nonvirtualized functional elements Aligned with parallel IT development OVS OVS SDN Controller SDN Forwarder OVS OVS
The NFV Concept Network functions are fully defined by SW, minimising dependence on HW constraints DPI CG-NAT BRAS Firewall PE Router GGSN/ SGSN VIRTUAL NETWORK FUNCTIONS FUNCTION COMMON HW (Servers & Switches) CAPACITY 9
The NFV Framework Logical Abstractions End Point E2E Network Service Network Service VNF VNF End Point Logical Links VNF VNF VNF SW Instances VNF : Virtualized Network Function VNF Instances VNF VNF VNF VNF NFV Infrastructure Virtual Resources Virtual Compute Virtual Storage Virtual Network Virtualization SW Virtualization Layer HW Resources Compute Storage Network 10
State of the Art: NFV and SDN
Actually Open collaboration in SDN/NFV has to go beyond pure open-source integration: Embrace collaborative development Clear current unknowns in the SDN+NFV equation SDN NFV
Outlook
Just Three Gears??? Leads to agility Reduces CAPEX and OPEX Network Functions Virtualisation Open Mind Software Defined Networking Creates abstractions to enable faster innovation Open Innovation Analyse and capture new, disruptive ideas Creates competitive supply of innovative applications by third parties 15
Intent Based Networking Currently SDN means programming how we do things Two vendors, two philosophies Three vendors, three philosophies So what about what we want to do? Currently router Boston_Router interface Eth0 ip address 10.0.0.1 mask 255.255.255.252 interfaces{ interface ge-0-1-1 { family inet { address 10.0.0.2/30; When all I want to say is: Link NY_Boston Type WAN EndNodes NY, Boston;
Intent Based Networking Projects Open Networking Foundation Through OpenSource SDN http://opensourcesdn.org/ossdn-projects/ Project Aspen Intent Based Northbound Interface API for real time media Project Boulder Intent Based Nothbound Interface OpenDaylight OpenDaylight Network Intent Composition https://github.com/opendaylight/nic IBNEMO: network modelling language http:nemo-project.net
Conclusion
Conclusion SDN and NFV are here to stay Help Network Operators create a more agile network at a reasonable cost We need to find the perfect mix of SDN and NFV We need an open mind to pick up other developments with big innovation potential INTENT as a way to simplify network operations further RINA to maximise reuse and simplify network structure