distributed route aggregation
play

Distributed Route Aggregation on the GlObal Network (DRAGON) Joo - PowerPoint PPT Presentation

Jan 19, 2024 •230 likes •778 views

Distributed Route Aggregation on the GlObal Network (DRAGON) Joo Lus Sobrinho 1 Laurent Vanbever 2 , Franck Le 3 , Jennifer Rexford 4 ACM CoNEXT 2014, Sydney 1 Instituto de Telecomunicaes, 1 IST Universidade de Lisboa 2 ETH Zurich, 3 IBM

  1. Distributed Route Aggregation on the GlObal Network (DRAGON) João Luís Sobrinho 1 Laurent Vanbever 2 , Franck Le 3 , Jennifer Rexford 4 ACM CoNEXT 2014, Sydney 1 Instituto de Telecomunicações, 1 IST Universidade de Lisboa 2 ETH Zurich, 3 IBM T. J. Watson Research, 4 Princeton University

  2. Last year in the news (August 2014) Some routers could not process the +512 K IPv4 prefixes they were learning about 2

  3. Not a scalable routing system 1.0.0.0/16 1.0.0.0/16 AS 2 1.0.0.0/16 AS 3 AS 1 1.0.0.0/16 origin AS 4 1.0.0.0/16 1.0.0.0/16 AS 6 AS 5 Most of the originated prefixes are routed globally (by BGP) 3

  4. Not a scalable routing system 1.0.0.0/16 1.0.0.0/16 1.1.0.0/16 AS 2 1.0.0.0/16 1.1.0.0/16 AS 3 AS 1 1.1.0.0/16 1.0.0.0/16 origin AS 4 1.1.0.0/16 1.0.0.0/16 1.0.0.0/16 AS 6 1.1.0.0/16 1.1.0.0/16 origin AS 5 Most of the originated prefixes are routed globally (by BGP) 4

  5. Not a scalable routing system 1.0.0.0/16 1.0.0.0/16 1.1.0.0/16 AS 2 1.0.0.0/16 1.1.0.0/16 AS 3 1.0.1.0/24 AS 1 1.1.0.0/16 1.0.1.0/24 1.0.1.0/24 1.0.0.0/16 origin AS 4 1.1.0.0/16 1.0.1.0/24 1.0.0.0/16 1.0.0.0/16 AS 6 1.1.0.0/16 1.1.0.0/16 origin AS 5 1.0.1.0/24 origin 1.0.1.0/24 Most of the originated prefixes are routed globally (by BGP) 5

  6. No scalability: poor performance • Forwarding tables (FIBs) growth & address look- up time increase • Routing tables (RIBs) growth • BGP session set-up time increase • Churn & convergence time increase 6

  7. Further scalability concerns • IPv6 prefixes can be formed in potentially larger numbers than IPv4 prefixes • Secure BGP adds computational overhead to routing processes 7

  8. DRAGON Distributed solution to scale the Internet routing system Basic DRAGON : 49% savings on routing state Full DRAGON : 79% savings on routing state No changes to the BGP protocol No changes to the forwarding plane Readily implemented with updated router software 8

  9. Outline • Scalability: global view • DRAGON: filtering strategy • DRAGON: aggregation strategy • DRAGON: performance evaluation • Conclusions 9

  10. Outline • Scalability: global view • DRAGON: filtering strategy • DRAGON: aggregation strategy • DRAGON: performance evaluation • Conclusions 10

  11. Scalability: global view (routing) Specificity Prefix q is more specific than prefix p if the bits of p are AS 1 1.0.0.0/16 the first bits of q 1.0.0.0/16 origin AS 2 1.0.1.0/24 1.0.0.0/16 AS 3 1.0.1.0/24 origin Propagation of more specific prefixes only in a small vicinity of their origin ASs 11

  12. Scalability: global view (forwarding) dest. addr. data-packet AS 1 1.0.0.0/16 1.0.1.1 1.0.0.0/16 origin AS 2 1.0.1.0/24 1.0.0.0/16 AS 3 1.0.1.0/24 origin Most ASs forward data-packets on the (aggregated) less specific prefixes 12

  13. Scalability: global view (forwarding) dest. addr. data-packet AS 1 1.0.0.0/16 1.0.0.0/16 origin AS 2 1.0.1.0/24 1.0.1.1 1.0.0.0/16 AS 3 1.0.1.0/24 origin 13

  14. Hope for scalability? Hierarchies provider AS 1 1.0.0.0/16 AS hierarchy Prefix hierarchy customer AS 2 1.0.1.0/24 AS-hierarchy aligned with prefix hierarchy 14

  15. Hope for scalability? Clustering Routing Information Registry (RIR) AS 3 1.0.0.0/24 AS 4 1.0.1.0/24 1.0.2.0/23 AS 5 1.0.0.0/24 AS 1 1.0.1.0/24 AS 6 AS 2 1.0.2.0/23 AS 3 1.0.0.0/24 + 1.0.1.0/24 + 1.0.2.0/23 = 1.0.0.0/22 Geography roughly clusters together ASs with aggregatable address space 15

  16. Challenge: global vs. local How to realize the global view through automated local routing decisions? especially, given that the Internet routing system is as decentralized as it can be: • each AS decides where to connect • each AS decides where to acquire address space • each AS sets its own routing policies 16

  17. Outline • Scalability: global view • DRAGON: filtering strategy • DRAGON: aggregation strategy • DRAGON: performance evaluation • Conclusions 17

  18. Filtering strategy • Locally filter the more specific prefixes when possible – no black holes – respect routing policies • Use built-in incentives to filter locally – save on forwarding state – forward along best route • Exchange routing information with standard BGP 18

  19. Providers, customers, and peers peer peer #1 #2 provider #3 #4 customer #5 #6 19

  20. Prefixes #1 #2 p : origin #3 #4 #5 #6 q : origin #6 originates q (1.0.0.0/24); #4 originates p (1.0.0.0/16) q more specific than p 20

  21. Routes Route #1 #2 Association between a prefix and an attribute, from a totally ordered set of attributes p : origin #3 #4 q -route #5 #6 q : origin (route pertaining to q ) 21

  22. Gao-Rexford routing policies #1 #2 route attributes: “ customer ” “ peer ” p : origin #3 #4 “ provider ” q -route #5 #6 q : origin preferences: customer then peer then provider exportations: all routes from customers; all routes to customers 22

  23. Gao-Rexford routing policies #1 #2 route attributes: “ customer ” q : cust. “ peer ” p : origin #3 #4 “ provider ” q : cust. q -route #5 #6 q : origin preferences: customer then peer then provider exportations: all routes from customers; all routes to customers 23

  24. Gao-Rexford routing policies #1 #2 route attributes: q : cust. “ customer ” q : cust. “ peer ” p : origin #3 #4 “ provider ” q : cust. q -route #5 #6 q : prov. q : origin preferences: customer then peer then provider exportations: all routes from customers; all routes to customers 24

  25. Gao-Rexford routing policies #1 #2 route attributes: q : cust. q : peer “ customer ” q : cust. “ peer ” p : origin #3 #4 “ provider ” q : cust. q -route #5 #6 q : prov. q : origin preferences: customer then peer then provider exportations: all routes from customers; all routes to customers 25

  26. Final state for prefix q #1 #2 route attributes: q : cust. q : peer “ customer ” q : cust. “ peer ” p : origin #3 #4 “ provider ” q : cust. #5 #6 q : prov. q : origin 26

  27. Final state for prefixes q and p p : cust. p : peer #1 #2 route attributes: q : cust. q : peer “ customer ” p : prov. q : cust. “ peer ” p : origin #3 #4 “ provider ” q : cust. p : prov. p : prov. #5 #6 q : prov. q : origin forwarding: longest prefix match rule 27

  28. Filtering code (FC) p : cust. p : peer Filtering Code (FC) #1 #2 q : cust. q : peer p : prov. Other than origin of p , q : cust. in the presence of p, p : origin filter q if only if: #3 #4 q : cust. attribute of p -route same or preferred to p : prov. p : prov. #5 #6 attribute of q -route q : prov. q : origin ASs that filter q upon execution of FC 28

  29. AS 2 applies FC p : cust. p : peer #1 #2 q : cust. filtered prefix p : prov. q : cust. p : origin #3 #4 q : cust. AS forgoes q p : prov. p : prov. #5 #6 q : prov. q : origin • AS 2 saves on forwarding state • AS 1 is oblivious of q ; it saves on AS 2 filters q forwarding and routing state 29

  30. All ASs apply FC p : cust. p : peer #1 #2 q : cust. filtered prefix p : prov. q : cust. p : origin #3 #4 q : cust. AS forgoes q p : prov. p : prov. #5 #6 q : prov. q : origin forwarding to q using less specific p AS 1, AS 2, and AS 3 forgo q 30

  31. Global property: correctness p : cust. p : peer #1 #2 q : cust. p : prov. q : cust. p : origin #3 #4 q : cust. forwarding data-packets with destination in q p : prov. p : prov. #5 #6 q : prov. q : origin Correctness : no routing anomalies (no black holes) 31

  32. Global property: route consistency p : cust. p : peer #1 #2 q : cust. p : prov. q : cust. p : origin #3 #4 q : cust. forwarding data-packets with destination in q p : prov. p : prov. #5 #6 q : prov. q : origin Route consistency : attribute of route used to forward data- packets is preserved Optimal route consistency : set of ASs that forgo q is maximal 32

  33. Partial deployment p : peer #1 q : cust. p : cust. #2 forwarding data-packets q : cust. with destination in q p : peer #3 p : cust. q : peer #4 q : cust. p : origin #5 q : cust. ASs that filter q upon execution of FC p : prov. #6 q : origin 33

  34. Partial deployment: incentives p : peer #1 q : cust. p : cust. #2 forwarding data-packets q : peer with destination in q p : peer #3 p : cust. q : prov. #4 q : cust. p : origin #5 q : cust. p : prov. #6 q : origin AS 2 (and AS 3) has a double incentive to apply the FC: • saves on forwarding state • improves attribute of route used to forward data-packets 34

  35. Partial deployment: incentives p : peer #1 q : cust. p : cust. #2 forwarding data-packets q : peer with destination in q p : peer #3 p : cust. q : prov. #4 q : cust. p : origin #5 q : cust. p : prov. #6 q : origin AS 2 applies FC AS 2 reverts to forwarding data-packets with address in q to AS 4 35

  36. Partial deployment: route consistency p : peer #1 q : cust. p : cust. #2 forwarding data-packets q : cust. with destination in q p : peer #3 p : cust. q : peer #4 q : cust. p : origin #5 q : cust. p : prov. #6 q : origin 36

  37. Partial deployment: route consistency p : peer #1 q : cust. p : cust. #2 forwarding data-packets q : cust. with destination in q p : peer #3 p : cust. q : peer #4 q : cust. p : origin #5 q : cust. p : prov. #6 q : origin First to apply FC are ASs that elect a peer or provider q -route 37

Recommend

R.A.I.D.F.S Randomized Aggregation Independent Distributed File System P2P Distributed File

R.A.I.D.F.S Randomized Aggregation Independent Distributed File System P2P Distributed File

R.A.I.D.F.S Randomized Aggregation Independent Distributed File System P2P Distributed File System with an API for Map-Reduce Integration Sven Reber, Jrmy Gotteland, David Froelicher, Alban Marguet, Pascal Cudr, Valrian Pittet Context

253 views • 21 slides
Multirobot autonomous landmine detection aggregation Simulation using distributed multisensor

Multirobot autonomous landmine detection aggregation Simulation using distributed multisensor

Motivation Problem Background Market-based Multirobot autonomous landmine detection aggregation Simulation using distributed multisensor information Results Conclusion aggregation Janyl Jumadinova, Raj Dasgupta C-MANTIC Research Group

511 views • 24 slides
Decision on expanding metering and telemetry options initiative (aggregation of distributed

Decision on expanding metering and telemetry options initiative (aggregation of distributed

Decision on expanding metering and telemetry options initiative (aggregation of distributed energy resources) Tom Flynn Infrastructure Policy Development Manager Board of Governors Meeting General Session July 16, 2015 ISO Confidential

942 views • 10 slides
Data Collection and Aggregation Data Collection and Aggregation 1 Challenges: data Challenges:

Data Collection and Aggregation Data Collection and Aggregation 1 Challenges: data Challenges:

Data Collection and Aggregation Data Collection and Aggregation 1 Challenges: data Challenges: data Data type: numerical sensor readings. Rich and massive data, spatially distributed and correlated. Data dynamics: data

714 views • 48 slides
Route 147 and Route 11 Roadway Reconstruction Project PA Route 147 Section 110 US Route 11

Route 147 and Route 11 Roadway Reconstruction Project PA Route 147 Section 110 US Route 11

Route 147 and Route 11 Roadway Reconstruction Project PA Route 147 Section 110 US Route 11 Section 113 Northumberland County LARSON DESIGN GROUP, INC. DEPARTMENT OF TRANSPORTATION 715 JORDAN AVENUE 1000 COMMERCE PARK DRIVE, SUITE 201

662 views • 16 slides
Distributed Aggregation for Data- Parallel Computing Interfaces and Implementations Yuan Yu

Distributed Aggregation for Data- Parallel Computing Interfaces and Implementations Yuan Yu

Distributed Aggregation for Data- Parallel Computing Interfaces and Implementations Yuan Yu Pradeep Kumar Gunda Michael Isard Microsoft Research Silicon Valley Dryad and DryadLINQ Automatic query plan generation by DryadLINQ Automatic

721 views • 30 slides
Visualizing Big Data Outliers through Distributed Aggregation Leland Wilkinson. Proc VAST 2017,

Visualizing Big Data Outliers through Distributed Aggregation Leland Wilkinson. Proc VAST 2017,

Visualizing Big Data Outliers through Distributed Aggregation Leland Wilkinson. Proc VAST 2017, TVCG to appear. Theodore Smith CPSC 547 Nov 7, 2017 Outliers General definition Observations which appear to be inconsistent with the

538 views • 17 slides
CSC 369: Distributed Computing Alex Dekhtyar April 20 Day 7: MongoDB Aggregation Pipeline, Part

CSC 369: Distributed Computing Alex Dekhtyar April 20 Day 7: MongoDB Aggregation Pipeline, Part

CSC 369: Distributed Computing Alex Dekhtyar April 20 Day 7: MongoDB Aggregation Pipeline, Part II Housekeeping: plan for the week April 20 Grading: Lecture: joins Lab 1: not certain Lab: practice Lab 2: end of the week

483 views • 34 slides
Skype Features: Distributed indexing service to route your incoming calls to wherever

Skype Features: Distributed indexing service to route your incoming calls to wherever

Skype Features: Distributed indexing service to route your incoming calls to wherever your are logged in. Conference calls Skype Me Connector services, e.g., jyve.com Challenges Session Initiation: NATs and

509 views • 4 slides
The Internet Route Registry and You: A Tier 1 Network Perspective Brian Foust Sr. Director,

The Internet Route Registry and You: A Tier 1 Network Perspective Brian Foust Sr. Director,

The Internet Route Registry and You: A Tier 1 Network Perspective Brian Foust Sr. Director, Customer Solutions NTT Communications Global IP Network AS2914 What is the Internet Route Registry? A distributed database of route and

417 views • 39 slides
Motivation Computation and Aggregation of Quantiles Application at Lucent Technologies: from

Motivation Computation and Aggregation of Quantiles Application at Lucent Technologies: from

Motivation Computation and Aggregation of Quantiles Application at Lucent Technologies: from Data Streams software to monitor distributed IP-based services. John Chambers, David James, Goal: characterize various metrics (e.g. e- Diane

392 views • 3 slides
Distributed in exclusivity by Circuit Paul Ricard 2760 route des hauts du Camp 83330 LE

Distributed in exclusivity by Circuit Paul Ricard 2760 route des hauts du Camp 83330 LE

Distributed in exclusivity by Circuit Paul Ricard 2760 route des hauts du Camp 83330 LE CASTELLET (France) ZAMAR THERAPY " Optimization of perioperative care with cryo-thermoregulation What is the Zamar Therapy ?

387 views • 18 slides
Elmwood Park: Electricity Aggregation Developing an Opt-In Municipal Aggregation Program to

Elmwood Park: Electricity Aggregation Developing an Opt-In Municipal Aggregation Program to

Elmwood Park: Electricity Aggregation Developing an Opt-In Municipal Aggregation Program to Serve Residents of Elmwood Park What is Electricity Aggregation? Purchasing authority for local governments The option for the Village to

226 views • 12 slides
simplifying the customer experience through account aggregation Sim Sangha Business Development

simplifying the customer experience through account aggregation Sim Sangha Business Development

simplifying the customer experience through account aggregation Sim Sangha Business Development Director total account aggregation James Born 29th June 2007 What is account aggregation? Why is account aggregation important to your

605 views • 11 slides
The Axiomatic Method in Social Choice Theory: Preference Aggregation, Judgment Aggregation, Graph

The Axiomatic Method in Social Choice Theory: Preference Aggregation, Judgment Aggregation, Graph

Axiomatic Method Lorentz Center, June 2015 The Axiomatic Method in Social Choice Theory: Preference Aggregation, Judgment Aggregation, Graph Aggregation Ulle Endriss Institute for Logic, Language and Computation University of Amsterdam Ulle

393 views • 20 slides
Aggregation Announcements Aggregation Aggregate Functions So far, all SQL expressions have

Aggregation Announcements Aggregation Aggregate Functions So far, all SQL expressions have

Aggregation Announcements Aggregation Aggregate Functions So far, all SQL expressions have referred to the values in a single row at a time [expression] as [name], [expression] as [name], ... select [columns] from [table] where [expression]

242 views • 10 slides
1 Three Dimensional Aggregation (con.t) Three Dimensional Aggregation (con.t) If we need to

1 Three Dimensional Aggregation (con.t) Three Dimensional Aggregation (con.t) If we need to

Data Cube: A Relational Aggregation Operator Topic Outline Generalizing Group-By, Cross-Tab, and Sub-Totals J. Gray, al, Microsoft Research F. Pellow, al, IBM Research Visualization and dimension reduction The relational representation of

367 views • 4 slides
Aggregation: A Brief Overview January 2011 () Aggregation January 2011 1 / 20 Macroeconomic

Aggregation: A Brief Overview January 2011 () Aggregation January 2011 1 / 20 Macroeconomic

Aggregation: A Brief Overview January 2011 () Aggregation January 2011 1 / 20 Macroeconomic Aggregates Consumption, investment, real GDP, labour productivity, TFP, physical capital, human capital (quantity and quality), price level,

448 views • 20 slides
U.S. Bike Route 1 City of Savannah / Sustainability Office/ U.S. Bike Route 1/ March 28, 2019

U.S. Bike Route 1 City of Savannah / Sustainability Office/ U.S. Bike Route 1/ March 28, 2019

U.S. Bike Route 1 City of Savannah / Sustainability Office/ U.S. Bike Route 1/ March 28, 2019 U.S. Bike Route 1 Background U.S. Bicycle Route System is a national network of public bike travel routes Over 13,000 miles are rideable

519 views • 8 slides
Mitigation of BGP Route Leaks ietf-idr-route-leak-detection-mitigation-06 (Route leak definition:

Mitigation of BGP Route Leaks ietf-idr-route-leak-detection-mitigation-06 (Route leak definition:

Methods for Prevention, Detection and Mitigation of BGP Route Leaks ietf-idr-route-leak-detection-mitigation-06 (Route leak definition: RFC 7908) K. Sriram, D. Montgomery, B. Dickson, K. Patel, and A. Robachevsky IDR Working Group Meeting,

244 views • 9 slides
Robust Aggregation in Sensor Robust Aggregation in Sensor Networks Networks Jie Gao Computer

Robust Aggregation in Sensor Robust Aggregation in Sensor Networks Networks Jie Gao Computer

Robust Aggregation in Sensor Robust Aggregation in Sensor Networks Networks Jie Gao Computer Science Department Stony Brook University 1 Papers Papers [Nath04] Suman Nath, Phillip B. Gibbons, Zachary Anderson, and Srinivasan Seshan,

616 views • 43 slides
Data Collection and Aggregation 1 Challenges: data Data type: numerical sensor readings.

Data Collection and Aggregation 1 Challenges: data Data type: numerical sensor readings.

Data Collection and Aggregation 1 Challenges: data Data type: numerical sensor readings. Rich and massive data, spatially distributed and correlated. Data dynamics: data streaming and aging. Uncertainty, noise, erroneous

1.45k views • 99 slides
RBridge Aggregation Mingui Zhang, Donald Eastlake zhangmingui@huawei.com RBridge Aggregation 1

RBridge Aggregation Mingui Zhang, Donald Eastlake zhangmingui@huawei.com RBridge Aggregation 1

RBridge Aggregation Mingui Zhang, Donald Eastlake zhangmingui@huawei.com RBridge Aggregation 1 Single AF: Loop Avoidance To avoid loops involving native frames, TRILL allows only a single Appointed Forwarder for one VLAN on a local link.

554 views • 13 slides
Route testing dialer Break through your data What is a dialer? The route testing dialer route

Route testing dialer Break through your data What is a dialer? The route testing dialer route

Route testing dialer Break through your data What is a dialer? The route testing dialer route testing dialer is a software module that lets you test quality of routes quality of routes provided by your vendors by automatically dialing

362 views • 10 slides

More recommend