Icccn 1.0
Download as PPTX, PDF0 likes5,008 views
Software is increasingly powering technological innovation and disrupting traditional industries as more aspects become software-defined. Mobile data usage continues rising rapidly, driven mainly by video. The internet economy is advancing towards a service-centric model. Software-defined networking aims to make networks more programmable through logical abstractions, enabling new capabilities like traffic engineering and dynamic configuration. However, distributed systems and concurrency present challenges to scaling that require ongoing research.
![Open Network Foundation
“The physical separation of the network control plane from the
forwarding plane, and where a control plane controls several
devices”[1]
Software Defined Networking?
1. https://www.opennetworking.org/sdn-resources/sdn-definition](https://image.slidesharecdn.com/icccn-1-130814100042-phpapp01/85/Icccn-1-0-5-320.jpg)
![The cost of deadlock free consensus
1+(N-1)+[N*(N-1)]+N+1
0
50
100
150
200
250
300
2 3 4 5 6 7 8 9 10 11 12 13 14 15
Messages
Messages
N = # of Acceptors
Acceptors/Learners
Note: Classic Multi-Paxos without optimization](https://image.slidesharecdn.com/icccn-1-130814100042-phpapp01/85/Icccn-1-0-20-320.jpg)
Icccn 1.0
- 1. Software is Eating the World Gary Berger Technical Leader, Cisco Systems Inc. gaberger@cisco.com Image: http://www.crunchzilla.com/code-monster
- 2. Mobile computing continues to dominate capacity 0 2 4 6 8 10 12 2011 2012 2013 2014 2015 2016 Mobile Data Traffic (Exabytes/Month) 0% 20% 40% 60% 80% 100% Mobile Data Transfer Distribution Other Web Video Source: Cisco VNI Mobile 2012 Source: ByteMobile Mobile Analytics Report 2012
- 3. Internet Economy Advancing towards Services • Time shared system • Explicit control • Restricted scope • Tightly Coupled • Vertically Integrated Database CentricClient Centric Service CentricHost Centric • Desktop applications • Centralized File & Print • Many dependencies • Low network utilization • Evolution of Client/Server • 4GL Programming • Stored Procedures • Vertically Integrated • Proprietary “Technical Debt” • Loosely coupled components • Web based interactions • Activities across multiple hosts • Global scope • Data driven constraints Web Centric • Normalized Presentation Layer • Activities scheduled off isolated hosts • Self-Described Data “New Economy” Sparse to Dense
- 4. Software Defined Everything? Software Defined Networking Software Defined Storage Software Defined Security Etc..
- 5. Open Network Foundation “The physical separation of the network control plane from the forwarding plane, and where a control plane controls several devices”[1] Software Defined Networking? 1. https://www.opennetworking.org/sdn-resources/sdn-definition
- 6. “A software-defined network can be flexible enough to avoid the application assumptions of designers of previous kinds of networks”. Derived from Staple, Werbach “The End Of Spectrum Scarcity” Software Defined Networking
- 7. Adoption
- 8. The “Power Wall”, 2004 Herb Sutter “Free Lunch is Over”, 2005 Economic crash, 2008 Data, Data, Data.. Many Contributing Factors
- 9. 1. Managing the network at scale! 2. Control access to network resources through declarative (policy) and compositional (predicate) based programming paradigms 3. Leverage the increasing network node degree (higher bi- sectional bandwidth) for performance and availability 4. Static -> dynamic configuration and reconfiguration 5. Enabling an architecture for Ubiquitous Computing 6. Everything and Anything at the moment! What is it about! What its mostly about: Leveraging Global Information to optimize for: • Traffic Engineering • Admission control • Isolation • Filtering and • Forwarding
- 10. OpenDayLight
- 11. Known problems have gone back 40 years. Quite possibly vendors have been listening to the wrong people? Many attempts at programming the network but the consumer demanded a weak and confusing interface (i.e. CLI, SNMP, NetCONF) which have polluted the design space We still don’t have a formal way of thinking about network architecture The Network Disruption
- 12. 5 Oct 201112 RINA Architecture As many layers as you need “Layers contain distributed state of a specific scope”
- 13. Geomorphic View of Classic Internet Application 1 Application 2 1 2 3 4 Gateway Gateway Internet Core LAN 1 LAN 2 LAN 3
- 14. NetAPI Programmatic Forwarding Secure communications Policy Driven Inter-domain routing (i.e. Pathlets) Framework for Internet Innovation
- 15. But really distributed computing is hard!! Concurrency Contention Coherency
- 16. Universal Scalability Law C(N) = N 1+a(N -1)+ bN(N -1) Contention (e.g. serialization, locking) Coherency (i.e. penalty incurred for maintaining consistency of shared writeable data) Source: A General Theory of Computational Scalability Based on Rational Functions, Gunther 2008 Concurrency
- 17. SDN Scaling still under active research How many controllers? Devices per controller? Partitioning strategy?
- 19. Deeper analysis shows we cannot just throw away peer communication • Data dissemination • Discovery • Availability
- 20. The cost of deadlock free consensus 1+(N-1)+[N*(N-1)]+N+1 0 50 100 150 200 250 300 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Messages Messages N = # of Acceptors Acceptors/Learners Note: Classic Multi-Paxos without optimization
- 21. SDN Continuum A spectrum which encompasses in-network state distribution at one extreme and complete separation of control and dataplane as the other extreme The purpose of which can result in either a discovery of an invariant protocol architecture which provides the foundation for the next generation Internet or paves the way for complete protocol customization.
- 22. Standards
- 23. IETF Bottleneck Count by Publication Status Total BEST CURRENT PRACTICE 213 DRAFT STANDARD 145 EXPERIMENTAL 417 HISTORIC 267 INFORMATIONAL 2164 INTERNET STANDARD 96 PROPOSED STANDARD 2597 UNKNOWN 906 Grand Total 6805
- 24. RFC By Year 0 50 100 150 200 250 300 350 400 450 500 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 2010 2013 Average of 265/yr over past decade
- 25. Gap is widening 0 50 100 150 200 250 300 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 2010 2013 BEST CURRENT PRACTICE DRAFT STANDARD EXPERIMENTAL HISTORIC INFORMATIONAL INTERNET STANDARD PROPOSED STANDARD UNKNOWN
- 26. Getting To Internet Standard? 1. There are at least two independent interoperating implementations with widespread deployment and successful operational experience. 2. There are no errata against the specification that would cause a new implementation to fail to interoperate with deployed ones. 3. There are no unused features in the specification that greatly increase implementation complexity. 4. If the technology required to implement the specification requires patented or otherwise controlled technology, then the set of implementations must demonstrate at least two independent, separate and successful uses of the licensing process. RFC 6410
- 27. “Network Virtualization is the abstraction of the control plane into layers”.. Scott Shenker Abstract the invariants from the set of complex control plane protocols Vertex Edge Property (tenant, vlan, min bw, min latency, open, closed, etc.) Network Virtualization
- 28. Address Virtualization Policy Virtualization Topology Virtualization Virtualization Profile
- 29. Address Coupling Application Node L2 (MAC) L3 (IP) Socket (local) Route Path Application Node L2 (MAC) L3 (IP) Socket (local) L2 and L3 address point to the interface not the node
- 30. Strict dependencies of identity and address force coupling of policy, telemetry and state.. Programming models still being thought about (Languages, Compilers, Runtimes) Virtualization provides the illusion of infinite resources but we must have “Mechanical Sympathy” (i.e. TCAM Space, SRAM, CPU Cycles) Encapsulation provides a means to decouple identity from location by adding a logical name space over a location dependent address space (i.e. TRILL, FabricPath, LISP, STT, VXLAN, NVGRE, etc..) Proper abstractions provide the invariant interfaces but do we have the right ones? Centralizing control provides a global view. Are we willing to throw away the local view (BFD, LAG, etc..)? Conclusion
- 31. Innovation at the Server NFV Enablement
- 33. Application Budgets BYOS (Bring Your Own Stack) 1400 881 341 69 3000 1962 881 338 4920 3259 1530 661 10Mpps 14.8Mpps(1x10GE) 29.6Mpps(2x10GE) 59.5Mpps(1x40GE) Intel E5-2600 @200 Cycles 8Core@2Ghz 16Core@2Ghz 16Core@3.2GHz Optimizations DPDK NETMAP PF_RING/LI BZERO
- 34. "The only way to get increased performance for new applications is for developers to be aware of new features in these chips. They have to be aware of what’s inside to make their code more efficient.” Krste Asanovic, UC Berkley Systems Driven Networking
- 35. SDN is a manifestation of over 20 years of technical debt which can no longer be swept under the rug IETF process is antiquated Crowd Sourcing + Parallel Programming + CMP allows for rapid prototyping Things are going to change.. “Network innovation is stifled by applying a limited set of design principals along with craft-like patching in a rigid architecture” Conclusion
Editor's Notes
- #5: Ate the time, people lived closer to one another, long distance calls were non-existent, costly and rare.
- #6: Ate the time, people lived closer to one another, long distance calls were non-existent, costly and rare.
- #10: 2. All web traffic should be forwarded to this load balancer, If web traffic is > x reroute to load balancer YUbiquitous Computing is anywhere, anytime, anydevice.IoT, M2M, etc..
- #20: LAGBFD
- #21: 6=4310=11114-211
- #22: Protocol Oblivious ForwardingPathlets
- #24: RFC 6410 Moves to a two-tier model
- #25: Average over past 20 years is 266RFC’s per Year
- #27: 2011
- #28: Tenant, Location, Fairness, Isolation
- #30: Routing on a path causes liveness issues.
- #31: Additional namespaces and state
- #33: Source: http://www.intel.com/content/dam/www/public/us/en/documents/solution-briefs/communications-packet-processing-brief.pdf
- #34: Intel® Xeon® processor E5-2658 (2 sockets) at 2.1 GHz, 20 MB L3 Cache, 90W; Intel® C604 chipset; 16 x 4 GB RDIMM DDR3-133 MHz Three I/O controllers per socket provide a total of 80 PCIe express lanes supporting latest gen. 3 technology of up to 8Gbps per lane.67.2ns interval40 PCIe * 8GHz = 320Gbps?