Network Aware Forward Caching Presenter: Alexandre Gerber Jeffrey Erman, Mohammad T. Hajiaghayi, Dan Pei, Oliver Spatscheck AT&T Labs Research April 24.

Slides:



Advertisements
Similar presentations
Computer Concepts – Illustrated 8th edition
Advertisements

Greening Backbone Networks Shutting Off Cables in Bundled Links Will Fisher, Martin Suchara, and Jennifer Rexford Princeton University.
Optimizing Cost and Performance for Multihoming Nick Feamster CS 6250 Fall 2011.
Multihoming and Multi-path Routing
Page 1 Approximately Maximum Bandwidth Routing for Slotted Wireless Ad Hoc Networks Approximately Maximum Bandwidth Routing for Slotted Wireless Ad Hoc.
UNITED NATIONS Shipment Details Report – January 2006.
Network Resource Broker for IPTV in Cloud Computing Lei Liang, Dan He University of Surrey, UK OGF 27, G2C Workshop 15 Oct 2009 Banff,
Extending Drive-Thru Data Access by Vehicle-to-Vehicle Relay Jing Zhao Todd Arnold Yang Zhang Guohong Cao Pennsylvania State University September 15 th,
All rights reserved © 2006, Alcatel Grid Standardization & ETSI (May 2006) B. Berde, Alcatel R & I.
Skynet: A Cloud-Hopping Data Transfer Architecture Aleksandar Kuzmanovic
and 6.855J Cycle Canceling Algorithm. 2 A minimum cost flow problem , $4 20, $1 20, $2 25, $2 25, $5 20, $6 30, $
Presentation | P2P Media Summit CacheLogic Advanced Solutions for P2P Networks Presentation by Andrew Parker, CTO
Introduction to PeerApp
1Abacast - Confidential1 Hybrid Content Delivery Network (CDN) Technologies and Services.
October 27, 2008 DCIA P2P and Video Conference Robert Levitan, CEO Can You Afford NOT to Use The Cloud?
1Abacast - Confidential1 Hybrid Content Delivery Network (CDN) Technologies and Services.
P2P Media Summit LA May 2009 Mitchell Edwards CFO & GC BitTorrent, Inc.
© Copyright IBM Corporation 2009 IBM Global Business Services Innovation driving strategic choices for P2P Saul Berman Global & Americas Leader, IBM Strategy.
Abacast - Confidential1 Hybrid Content Delivery Network (CDN) Technologies and Services.
The Evolution of P2P Technology Robert Levitan, CEO Pando Networks P2P MEDIA SUMMIT, CES 2008.
P2P Market Conference March 2009 Mitchell Edwards CFO & GC BitTorrent, Inc.
1Abacast - Confidential1 Hybrid Content Delivery Network (CDN) Technologies and Services.
1Abacast - Confidential1 Hybrid Content Delivery Network (CDN) Technologies and Services.
0 - 0.
MULTIPLYING MONOMIALS TIMES POLYNOMIALS (DISTRIBUTIVE PROPERTY)
£1 Million £500,000 £250,000 £125,000 £64,000 £32,000 £16,000 £8,000 £4,000 £2,000 £1,000 £500 £300 £200 £100 Welcome.
Xia Zhou*, Stratis Ioannidis ♯, and Laurent Massoulié + * University of California, Santa Barbara ♯ Technicolor Research Lab, Palo Alto + Technicolor Research.
Chador Wangdi Page 1of15. Analog (up to 56k/ dial up access) Using a modem connected to your PC, users connect to the internet when the computer dials.
1 Jin Li Microsoft Research. Outline The Upcoming Video Tidal Wave Internet Infrastructure: Data Center/CDN/P2P P2P in Microsoft Locality aware P2P Conclusions.
Copyright © 2008 Roger Webster, Ph.D. EDW647 Internet For Educators Conclusion Roger W. Webster, Ph.D. Department of Computer Science Millersville University.
Cache Storage For the Next Billion Students: Anirudh Badam, Sunghwan Ihm Research Scientist: KyoungSoo Park Presenter: Vivek Pai Collaborator: Larry Peterson.
The IP Revolution. Page 2 The IP Revolution IP Revolution Why now? The 3 Pillars of the IP Revolution How IP changes everything.
Chapter 1: Introduction to Scaling Networks
IS THERE A CASE FOR MOBILE PHONE CONTENT PRE-STAGING? Santa Barbara, December 9-12, 2013 Alessandro Finamore Marco Mellia Zafar Gilani Konstantina Papagiannaki.
1 Sizing the Streaming Media Cluster Solution for a Given Workload Lucy Cherkasova and Wenting Tang HPLabs.
Copyright © Chang Gung University. Permission required for reproduction or display. On Femto Deployment Architecture and Macrocell Offloading Benefits.
Communications Research Centre (CRC) Defence R&D Canada – Ottawa 1 Properties of Mobile Tactical Radio Networks on VHF Bands Li Li & Phil Vigneron Communications.
© 2005 AT&T, All Rights Reserved. 11 July 2005 AT&T Enhanced VPN Services Performance Reporting and Web Tools Presenter : Sam Levine x111.
McGraw-Hill/Irwin Copyright © 2007 by The McGraw-Hill Companies, Inc. All rights reserved. Chapter Eleven Cost Behavior, Operating Leverage, and CVP Analysis.
IP Multicast Information management 2 Groep T Leuven – Information department 2/14 Agenda •Why IP Multicast ? •Multicast fundamentals •Intradomain.
Flow Aware Networking © 2007 Katedra Telekomunikacji AGH Flow Aware Networking Router model lead by prof. dr hab. inż. Andrzej Jajszczyk.
Created by Susan Neal $100 Fractions Addition Fractions Subtraction Fractions Multiplication Fractions Division General $200 $300 $400 $500 $100 $200.
RED-PD: RED with Preferential Dropping Ratul Mahajan Sally Floyd David Wetherall.
Getting Familiar with Web Pages 1 2 The Internet Worldwide collection of interconnected computer networks that enables businesses, organizations, governments,
Chapter 5 Test Review Sections 5-1 through 5-4.
Page 1 / 18 Internet Traffic Monitor IM Page 2 / 18 Outline Product Overview Product Features Product Application Web UI.
PSIRP Publish-Subscribe Internet Routing Paradigm 08-Oct /27.
Addition 1’s to 20.
Model and Relationships 6 M 1 M M M M M M M M M M M M M M M M
25 seconds left…...
Abhigyan, Aditya Mishra, Vikas Kumar, Arun Venkataramani University of Massachusetts Amherst 1.
Bytindex – Data Optimization Concept Technology briefing
A SMALL TRUTH TO MAKE LIFE 100%
Multimedia over DSL By Phil Moy. May 14, Agenda n DSL Forum Working Text 80 - Multiservice Architecture & Framework Requirements n DSL Forum Working.
19 Historical overview Main challenge: How to distribute content in high quality over the Internet cost-effectively? • Traditional “Best-effort” model:
Mohamed Hefeeda 1 School of Computing Science Simon Fraser University, Canada ISP-Friendly Peer Matching without ISP Collaboration Mohamed Hefeeda (Joint.
An Analysis of Internet Content Delivery Systems Stefan Saroiu, Krishna P. Gommadi, Richard J. Dunn, Steven D. Gribble, and Henry M. Levy Proceedings of.
Provisioning Content Distribution Networks for Streaming Media Jussara M. Almeida Derek L. Eager Michael Ferris Mary K. Vernon University of Wisconsin-Madison.
Tradeoffs in CDN Designs for Throughput Oriented Traffic Minlan Yu University of Southern California 1 Joint work with Wenjie Jiang, Haoyuan Li, and Ion.
© 2009 AT&T Intellectual Property. All rights reserved. Multimedia content growth: From IP networks to Medianets Cisco-IEEE ComSoc Webinar. Sept. 23, 2009.
1 Proceeding the Second Exercises on Computer and Systems Engineering Professor OKAMURA Laboratory. Othman Othman M.M.
Chapter 4. After completion of this chapter, you should be able to: Explain “what is the Internet? And how we connect to the Internet using an ISP. Explain.
By Sylvia Ratnasamy, Andrey Ermolinskiy, Scott Shenker Presented by Fei Jia Revisiting IP Multicast.
Othman Othman M.M., Koji Okamura Kyushu University 1.
An Efficient Approach for Content Delivery in Overlay Networks Mohammad Malli Chadi Barakat, Walid Dabbous Planete Project To appear in proceedings of.
Economic Incentives in Information- Centric Networking: Implications for Protocol Design and Public Policy Group Members: Muhammad Kamran Siddique Adel.
CONTENT DELIVERY NETWORKS
Content Delivery Networks: Status and Trends Speaker: Shao-Fen Chou Advisor: Dr. Ho-Ting Wu 5/8/
WHAT'S THE DIFFERENCE BETWEEN A WEB APPLICATION STREAMING NETWORK AND A CDN? INSTART LOGIC.
EE 122: Lecture 22 (Overlay Networks)
Presentation transcript:

Network Aware Forward Caching Presenter: Alexandre Gerber Jeffrey Erman, Mohammad T. Hajiaghayi, Dan Pei, Oliver Spatscheck AT&T Labs Research April 24 th, 2009

Outline Whats happening on the Internet? Traffic characterization Efficiency of existing delivery mechanisms Cacheability of HTTP traffic Lets revisit forward caching with a new twist! Network Aware Forward caching Page 2 © 2009 AT&T Intellectual Property. All rights reserved.

Data available to understand Internet traffic Application Mix and HTTP requests ( ): 100,000 broadband subscribers: US only: California, Texas, Illinois DSL not Cable Monitoring at the edge: BRAS (aggregation point) Application classification based on application header (e.g. MIME type) Efficiency of delivery mechanisms (October 2008): Complete Netflow based PoP to PoP Traffic matrix of a US broadband ISP Air Miles between Points of Presence (PoP) © 2009 AT&T Intellectual Property. All rights reserved. Page 3

Aggregate downstream traffic growth per subscriber Stable over the last 8 years in aggregate! 20-30% per subscriber per year Page 4 © 2009 AT&T Intellectual Property. All rights reserved.

Lets drill down: Application Mix over the last 2 years HTTP is back and it is growing fast! Thanks in part to Multimedia streams over HTTP! © 2009 AT&T Intellectual Property. All rights reserved. Page 5 Growing 3 times faster than the historical aggregate growth rate

Application Mix: January % of potentially reusable content © 2009 AT&T Intellectual Property. All rights reserved. Page 6 DownstreamUpstream Busy Hour Average

Lets keep drilling down: HTTP & Multimedia Its no longer about text and images, HTTP is the workhorse for data delivery (streaming or direct downloads) © 2009 AT&T Intellectual Property. All rights reserved. Page 7 /http/download accounts for 10.2% traffic of overall traffic while FTP accounts for 0.3% http accounts for 80% of the multimedia content delivery (flash)

Is this content delivered efficiently? CDNs are doing a good job! P2P protocols are not! © 2009 AT&T Intellectual Property. All rights reserved. Page 8 CDN traffic is traversing 60% fewer air miles than the average traffic P2P flows are traversing more air miles than the average traffic

Are CDNs big enough to have an impact? Yes, because they carry a significant fraction of large files. © 2009 AT&T Intellectual Property. All rights reserved. Page 9 46% of very large files are distributed by the 3 large CDNs studied

Are all the Points of Presence equal in the US? No, some are much closer/farther from content Page 10 © 2009 AT&T Intellectual Property. All rights reserved.

Are all the Autonomous Systems traversing the same distance on an ISPs backbone? No, some are much closer/farther from content Page 11 © 2009 AT&T Intellectual Property. All rights reserved.

What have we learned so far? 89% of the content during the busy hour may be reusable (HTTP content, Multimedia streams, P2P / File sharing) 68% of the content during the busy hour is coming from HTTP and HTTP is growing fast: it is the workhorse for data delivery and includes the strong growth of Multimedia streams There are big differences in the delivery of content when comparing air miles on an ISPs backbone: CDNs are doing a good job but other web content providers and P2P protocols are not Distance traversed to reach PoPs varies significantly What if we only cached the HTTP traffic that has to traverse the longest distance? © 2009 AT&T Intellectual Property. All rights reserved. Page 12

Lets first understand if HTTP content is really cacheable: Yes, it is! 32% of HTTP bytes served from the cache (24% of total downstream traffic during the busy hour) Cache size is reasonable (TB) Page 13 © 2009 AT&T Intellectual Property. All rights reserved. Based on 20,000 subscribers in January 2008

Network Aware Forward caching HTTP forward cache that is aware of the network: Understand the cost of each bit: Backbone cost: Air Miles traversed on ISPs network by each bit Source of traffic: free peering vs. paid transit traffic Lets only cache the bits that need it most Tradeoff between bandwidth cost vs. caching cost Caching decision made for: Each PoP: no caches? X caches? And each e ntity in each PoP: BGP prefix, AS, IP addresses, etc. Tradeoff between amount of traffic generated vs. storage requirement Combinatorial, knapsack like, NP-hard problem Can be solved with dynamic programming in pseudo polynomial time Or greedy heuristic by sorting entities in decreasing order of benefits © 2009 AT&T Intellectual Property. All rights reserved. Page 14

Solution sensitive to backbone costs and caching costs Page 15 © 2009 AT&T Intellectual Property. All rights reserved. Base scenario: $4/Mbps/Month for transit and backbone costs, $20K for caches (400 Mbps, 4TB) Sensitivity to Backbone costsSensitivity to Caching costs The optimal deployment is to cache 68% of the traffic. This is 37% better than a simple cache all solution!

Drilling down into the optimal solution: 25% of the PoPs dont need it a cache Only for 15% of the PoPs does it make sense to cache all the cacheable content Page 16 © 2009 AT&T Intellectual Property. All rights reserved.

Conclusion & Future Work Contribution: Characterization of Internet traffic & its distribution efficiency: There is a strong growth of HTTP content, especially multimedia streams Significant differences in distribution efficiencies Proposed and evaluated Network Aware Forward Caching for HTTP Future work: Explore other mechanisms to better distribute Internet traffic: network aware P2P, Multicast, Anycast, etc. Extend study to other environments: wireless IP networks © 2009 AT&T Intellectual Property. All rights reserved. Page 17