Network Protocols Designed for Optimizability Jennifer Rexford Princeton University

Slides:



Advertisements
Similar presentations
Jennifer Rexford Princeton University MW 11:00am-12:20pm Logically-Centralized Control COS 597E: Software Defined Networking.
Advertisements

Logically Centralized Control Class 2. Types of Networks ISP Networks – Entity only owns the switches – Throughput: 100GB-10TB – Heterogeneous devices:
Nanxi Kang Princeton University
1 BGP Anomaly Detection in an ISP Jian Wu (U. Michigan) Z. Morley Mao (U. Michigan) Jennifer Rexford (Princeton) Jia Wang (AT&T Labs)
XCP: Congestion Control for High Bandwidth-Delay Product Network Dina Katabi, Mark Handley and Charlie Rohrs Presented by Ao-Jan Su.
TIE Breaking: Tunable Interdomain Egress Selection Renata Teixeira Laboratoire d’Informatique de Paris 6 Université Pierre et Marie Curie with Tim Griffin.
1 Finding a Needle in a Haystack: Pinpointing Significant BGP Routing Changes in an IP Network Jian Wu (University of Michigan) Z. Morley Mao (University.
Traffic Engineering With Traditional IP Routing Protocols
1 Route Control Platform Making the Network Act Like One Big Router Jennifer Rexford Princeton University
1 Adapting Routing to the Traffic COS 461: Computer Networks Spring 2006 (MW 1:30-2:50 in Friend 109) Jennifer Rexford Teaching Assistant: Mike Wawrzoniak.
Traffic Engineering Jennifer Rexford Advanced Computer Networks Tuesdays/Thursdays 1:30pm-2:50pm.
1 Controlling High Bandwidth Aggregates in the Network.
Traffic Engineering for ISP Networks Jennifer Rexford Computer Science Department Princeton University
1 Traffic Engineering for ISP Networks Jennifer Rexford IP Network Management and Performance AT&T Labs - Research; Florham Park, NJ
Traffic Engineering in IP Networks Jennifer Rexford Computer Science Department Princeton University; Princeton, NJ
Traffic Engineering for ISP Networks Jennifer Rexford Computer Science Department Princeton University
Rethinking Traffic Management: Using Optimization Decomposition to Derive New Architectures Jennifer Rexford Princeton University Jiayue He, Ma’ayan Bresler,
Traffic Engineering for ISP Networks
A Routing Control Platform for Managing IP Networks Jennifer Rexford Computer Science Department Princeton University
Traffic Engineering for ISP Networks Jennifer Rexford Computer Science Department Princeton University
Design for Network Managability Mung Chiang and Jennifer Rexford Princeton University March 2007.
Internet Routing (COS 598A) Today: Intradomain Traffic Engineering Jennifer Rexford Tuesdays/Thursdays.
Wresting Control from BGP: Scalable Fine-grained Route Control UCSD / AT&T Research Usenix —June 22, 2007 Dan Pei, Tom Scholl, Aman Shaikh, Alex C. Snoeren,
Traffic Engineering for ISP Networks Jennifer Rexford Internet and Networking Systems AT&T Labs - Research; Florham Park, NJ
Internet Routing (COS 598A) Today: Interdomain Traffic Engineering Jennifer Rexford Tuesdays/Thursdays.
Traffic Engineering for ISP Networks Jennifer Rexford Computer Science Department Princeton University
Routing problems are easy to cause, and hard to diagnose (“Happy operators make happy packets”) Jennifer Rexford AT&T Labs—Research
1 Design and implementation of a Routing Control Platform Matthew Caesar, Donald Caldwell, Nick Feamster, Jennifer Rexford, Aman Shaikh, Jacobus van der.
Internet Routing (COS 598A) Today: Hot-Potato Routing Jennifer Rexford Tuesdays/Thursdays 11:00am-12:20pm.
A Routing Control Platform for Managing IP Networks Jennifer Rexford Princeton University
Measurement and Monitoring Nick Feamster Georgia Tech.
Dynamic routing – QoS routing Load sensitive routing QoS routing.
Rethinking Internet Traffic Management: From Multiple Decompositions to a Practical Protocol Jiayue He Princeton University Joint work with Martin Suchara,
Network Monitoring for Internet Traffic Engineering Jennifer Rexford AT&T Labs – Research Florham Park, NJ 07932
A Routing Control Platform for Managing IP Networks Jennifer Rexford Princeton University
1 Traffic Engineering for ISP Networks Jennifer Rexford IP Network Management and Performance AT&T Labs - Research; Florham Park, NJ
A Routing Control Platform for Managing IP Networks Jennifer Rexford Princeton University
Multipath Protocol for Delay-Sensitive Traffic Jennifer Rexford Princeton University Joint work with Umar Javed, Martin Suchara, and Jiayue He
1 Experimental Network Management Jennifer Rexford Princeton University
1 Network-wide Decision Making: Toward a Wafer-thin Control Plane Jennifer Rexford, Albert Greenberg, Gisli Hjalmtysson ATT Labs Research David A. Maltz,
Jennifer Rexford Princeton University MW 11:00am-12:20pm Wide-Area Traffic Management COS 597E: Software Defined Networking.
Transit Traffic Engineering Nick Feamster CS 6250: Computer Networks Fall 2011.
Network Sensitivity to Hot-Potato Disruptions Renata Teixeira (UC San Diego) with Aman Shaikh (AT&T), Tim Griffin(Intel),
DaVinci: Dynamically Adaptive Virtual Networks for a Customized Internet Jennifer Rexford Princeton University With Jiayue He, Rui Zhang-Shen, Ying Li,
Authors Renata Teixeira, Aman Shaikh and Jennifer Rexford(AT&T), Tim Griffin(Intel) Presenter : Farrukh Shahzad.
Traffic Engineering for ISP Networks Jennifer Rexford Internet and Networking Systems AT&T Labs - Research; Florham Park, NJ
Using Measurement Data to Construct a Network-Wide View Jennifer Rexford AT&T Labs—Research Florham Park, NJ
Jennifer Rexford Fall 2014 (TTh 3:00-4:20 in CS 105) COS 561: Advanced Computer Networks BGP.
9.7 Other Congestion Related Issues Outline Queuing Discipline Avoiding Congestion.
Jennifer Rexford Princeton University MW 11:00am-12:20pm Measurement COS 597E: Software Defined Networking.
DaVinci: Dynamically Adaptive Virtual Networks for a Customized Internet Jiayue He, Rui Zhang-Shen, Ying Li, Cheng-Yen Lee, Jennifer Rexford, and Mung.
Intradomain Traffic Engineering By Behzad Akbari These slides are based in part upon slides of J. Rexford (Princeton university)
Jennifer Rexford Fall 2010 (TTh 1:30-2:50 in COS 302) COS 561: Advanced Computer Networks Backbone.
Evolving Toward a Self-Managing Network Jennifer Rexford Princeton University
Evolving Toward a Self-Managing Network Jennifer Rexford Princeton University
P4 Amore! ( Or, How I Learned to Stop Worrying and Love P4) Jennifer Rexford Princeton University.
Mike Freedman Fall 2012 COS 561: Advanced Computer Networks Traffic Engineering.
The New Policy for Enterprise Networking Robert Bays Chief Scientist June 2002.
Network management IEX8175 RF Electronics
Separating Routing From Routers Jennifer Rexford Princeton University
Simulation and Exploration of
Jian Wu (University of Michigan)
COS 561: Advanced Computer Networks
6.829 Lecture 13: Software Defined Networking
Interdomain Traffic Engineering with BGP
COS 561: Advanced Computer Networks
Congestion Control, Quality of Service, & Internetworking
COS 561: Advanced Computer Networks
Backbone Traffic Engineering
BGP Instability Jennifer Rexford
Presentation transcript:

Network Protocols Designed for Optimizability Jennifer Rexford Princeton University

2 Measure, Model, and Control Topology/ Configuration Offered traffic Changes to the network Operational network Models, tools, scripts, databases measure control Network Management Dials Knobs

3 Knobs and Dials Knobs: configurable parameters –Buffering: Random Early Detection parameters –Link scheduling: weighted fair queuing weights –Path selection: link weights and routing policies Dials: measurement data –Traffic: link utilization, Netflow records, … –Performance: ping, download times, … –Routing: routing-protocol messages, tables, … Network management: read the dials and tune the knobs

4 Two Directions We Could Go Algorithms for setting knobs based on dials –E.g., setting RED parameters based on link load –E.g., setting link weights based on traffic matrix –E.g., setting access-control lists to block attacks Designing better knobs and dials –Maybe we can’t add all that much meaningful abstraction on top of what we’ve got underneath –Maybe we should design new protocols and mechanisms with optimization in mind –“Doing well in a class is much easier when you get to write the exam.” – Mung Chiang

5 Problem #1: No Algorithm For Setting the Knobs Random Early Detection (RED) –Several tunable parameters –Min and max thresholds on queue length, max probability, queue weight Average Queue Length Probability

6 Problem #1: RED Example Continued Settings have a big influence on performance –Good settings can improve the network “goodput” –Bad settings may offer no improvement, or (in some cases), worse performance No algorithm for optimizing the parameters –Settings based on general guidelines –Makes it difficult for operators to enable RED We need mechanisms that have algorithms for setting knobs.

7 Problem #2: Poor Dials to Guide Knob Settings Example: Random Early Detection –Appropriate parameters depend on many factors Number of active flows, flow durations, flow RTTs, … –Not easily measurable today on high-speed links We need measurements that support network management. Average Queue Length Probability

8 Problem #2: Poor Dials to Guide Knob Settings Example: Traffic engineering –Depends on knowing the traffic matrix M ij –Challenging to measure Resorting to inference of the traffic matrix Aggregating and joining lots of fine-grain data We need measurements that support network management. i j

9 Problem #3: Intractable Optimization Problems Example: Traffic engineering –Tuning link weights to the prevailing traffic –Leads to an NP-hard optimization problem –… forcing the use of local-search techniques We need protocols designed with knob optimization in mind.

10 Problem #4: Non-Linearities in the System Example: Hot-potato routing –Small change causes a big effect Failure, planned maintenance, or traffic engineering Routes to thousands of destinations shift at once … causing large shifts in traffic and many BGP updates SFO Dallas NYC ISP network dst We need protocols that make small reactions to small changes.

11 Design for Optimizability Creating protocols and mechanisms where –We know the algorithms for tuning the knobs –We have the measurements the algorithms need –The resulting optimization problems are tractable –The system does not have non-linearities Example approaches –Randomization –Increasing the degrees of freedom –Logically centralized control

12 Randomization Example: traffic engineering –Forward traffic in inverse proportion to path costs –… rather than using only the shortest paths –Leads to polynomial-time optimization problems

13 Increasing Degrees of Freedom Example: egress selection –Forward traffic to lowest ranked egress point –… as weighted sum of constant and path cost –E.g., keep using SFO even when cost goes to 11 –Enables integer programming solutions for tuning SFO Dallas NYC ISP network dst

14 Logically Centralized Control Example: Routing Control Platform (RCP) –Separate topology discovery from path selection –Collect topology and traffic data at servers –Apply optimization techniques for selecting routes –… and tell routers what forwarding tables to use RCP

15 Conclusions Protocols induce optimization problems –Read the dials and tune the knobs –Controls how the system performs Yet, optimization problems are often hard –Lack of predictive models –Missing measurement data –Computational intractability –Non-linearities in the system Design protocols with optimization in mind –Randomize, add degrees of freedom, decompose