1.End to end arguments in system design (1981) 2.Tussles in cyberspace: Defining Tomorrow’s Internet (2005) Nick McKeown CS244 Lecture 3 Architecture and Principles

2 End-to-End Arguments in System Design [Saltzer, Reed, Clark 1981] End-to-end in a nutshell “The function in question can completely and correctly be implemented only with the knowledge and help of the application standing at the end points of the communication system. Therefore, providing that questioned function as a feature of the communication system itself is not possible. (Sometimes an incomplete version of the function provided by the communication system may be useful as a performance enhancement.)”

3 Commonly used examples 1.Error handling in file transfer 2.End-to-end, versus in-network encryption 3.The partition between TCP and IP of error handling, flow control and congestion control.

What you said “For some applications, such as banking, application specific acknowledgements or transactions may be necessary to ensure consistency. In this respect, I think the question of what we define as an acceptable end-point is crucial in terms of what services we can expect the communication system to provide.” 4

What you said “I feel like this [paper] leads people to say, ‘This is bad because it violates the End to End Principle’, rather than, for example, ‘This is bad because it makes assumptions about what future applications might want.’ ” -- Jason Clavelli 5

6 Some consequences In layered design, the E2E principle provides guidance on where functions belong. “Dumb, minimal” network and “intelligent” end- points. Many argue that: E2E principle allowed the Internet to grow rapidly because innovation took place at the edge, in applications and services.

7 On the other hand… E2E principle appears to have become diluted: NATs, firewalls, VPN tunnel endpoints, … –Perhaps not surprising: E2E principle grew in an era of trust among users. Now network must protect itself. The network is no longer “dumb, minimal” –Now over 7,000 RFCs. –Router OS’s based on 100M lines of source code. Q: Is this a problem?

8 What belongs in, what out? Questions: Does routing belong in the “dumb, minimal” network? How about multicast, mobility, QoS…? Are NATs necessary, good, or evil? Is the E2E principle constraining innovation of the infrastructure?

9 Additional references [rfc3724] “The Rise of the Middle and the Future of End-to-End: Reflections on the Evolution of the Internet Architecture” - Kempf et al. [Blumenthal] “Rethinking the design of the Internet: The end-to-end arguments vs. the brave new world”, ACM Transactions on Internet Technology, Vol. 1, No. 1, August 2001, pp

Tussle in Cyberspace: Defining Tomorrow’s Internet 10 Actor-Network Theory (ANT) Assumes equal treatment of humans and non-humans in an interacting network. Distinction between ‘mediators’ and ‘intermediaries’: “silk and nylon”.

Context Why did the authors write the paper? What had changed since the Internet was invented? 11

Problem Statement “The Internet was created in simpler times. Its creators and early users shared a common goal —they wanted to build a network infrastructure to hook all the computers in the world together so that as yet unknown applications could be invented to run there. All the players, whether designers, users or operators, shared a consistent vision and a common sense of purpose.” “Perhaps the most important consequence of the Internet’s success is that the common purpose that launched and nurtured it no longer prevails.” 12

Types of Tussle 1.Economics 2.Trust 1.Openness 13

Trust “One of the most profound and irreversible changes in the Internet is that by and large, many of the users do not trust each other. The users of the Internet no longer represent a single community with common motivation and shared trust.” “[There] is a profound tussle, between people who want to be left alone, and people who want to bother them.” Q: How are we encouraged to think about this? 14

What you said “The design of interfaces is the common theme linking the Tussle paper and the end-to-end argument paper we are discussing…I agree the end-to-end argument is still valid today, but the transparency it encourages needs more careful considerations in a playing field where trust cannot be taken for granted.” -- Ana Klimovic 15

What you said “I'm pessimistic that there is much that system designers can do to bias the tussle in favor of the user at this point. Consider the recent net neutrality debate…” --Wen-Chen Chen 16

Questioning sacred cows 1.End to end argument Q: How is it affected by “tussles”? 1.Separate policy from mechanism Q: What does it mean? If the goal is to hook computers together and let users run any application they want, then a simple transparent network enables “user empowerment”, choice and innovation. 17

A lesson Hypothesis about QoS: Internet providers had no incentive to deploy. 1.There is a real cost to deploy 2.Users had no way to choose providers (local or remote). Q: How is this related to tussles? 18

Consequences What do the authors recommend we do? What are the concrete steps? 19

Context for Next Class 20

Software Defined Network (SDN) Packet Forwarding Packet Forwarding Packet Forwarding Packet Forwarding Packet Forwarding Packet Forwarding Packet Forwarding Packet Forwarding Packet Forwarding Packet Forwarding Control Global Network Map Control Plane Control Program Control Program Control Program

Software Defined Network A network in which the control plane is physically separate from the forwarding plane. and A single control plane controls several forwarding devices. (That’s it)

Software Control Plane Intended consequences... 1.Put network owners and operators in control. 2.Networks that are more reliable and more secure. 3.Networks that cost less: simpler, streamlined hardware. 4.Networks that cost less to operate (fewer features). SDN

An example Routing

function Dijkstra(Graph, source): for each vertex v in Graph: dist[v] := infinity ; previous[v] := undefined; dist[source] := 0 ; Q := the set of all nodes in Graph ; while Q is not empty: // The main loop u := vertex in Q with smallest distance in dist[] ; remove u from Q ; if dist[u] = infinity: break ; for each neighbor v of u: alt := dist[u] + dist_between(u, v) ; if alt < dist[v]: dist[v] := alt ; previous[v] := u ; decrease-key v in Q; return dist[], previous[]; end function Edsger Dikjstra Photo: Hamilton Richards

The Opte Project

1 2 3 “If, send to 3” Data “If a packet is going to B, then send it to output 3” 1.Figure out which routers and links are present. 2.Run Dijkstra’s algorithm to find shortest paths.

95% 5% 50,000 lines of code 1.Figure out which routers and links are present. 2.Run Dijkstra’s algorithm to find shortest paths.

Dijkstra Network OS IS-IS BGP MPLS Firewall… Global Network Map