Overhaul: Extending HTTP to Combat Flash Crowds Jay A. Patel & Indranil Gupta Distributed Protocols Research Group Department of Computer Science University.

Slides:



Advertisements
Similar presentations
Digital Library Service – An overview Introduction System Architecture Components and their functionalities Experimental Results.
Advertisements

CSE300-1 Profs. Steven A. Demurjian Q. Jin, J. Nam, Z. Qian and C. Phillips Computer Science & Engineering Department 191 Auditorium Road, Box U-155 The.
Secure Content Delivery in Information-Centric Networks: Design, Implementation, and Analyses Computer Science Department New Mexico State University,
Clayton Sullivan PEER-TO-PEER NETWORKS. INTRODUCTION What is a Peer-To-Peer Network A Peer Application Overlay Network Network Architecture and System.
Esma Yildirim Department of Computer Engineering Fatih University Istanbul, Turkey DATACLOUD 2013.
Using DSVM to Implement a Distributed File System Ramon Lawrence Dept. of Computer Science
1 Network Measurements of a Wireless Classroom Network Carey Williamson Nuha Kamaluddeen Department of Computer Science University of Calgary.
Small-world Overlay P2P Network
Toolbox Mirror -Overview Effective Distributed Learning.
Notes to the presenter. I would like to thank Jim Waldo, Jon Bostrom, and Dennis Govoni. They helped me put this presentation together for the field.
Web Caching Schemes1 A Survey of Web Caching Schemes for the Internet Jia Wang.
Rheeve: A Plug-n-Play Peer- to-Peer Computing Platform Wang-kee Poon and Jiannong Cao Department of Computing, The Hong Kong Polytechnic University ICDCSW.
Peer-to-peer Multimedia Streaming and Caching Service Jie WEI, Zhen MA May. 29.
Copyright © 2005 Department of Computer Science CPSC 641 Winter WAN Traffic Measurements There have been several studies of wide area network traffic.
An Analysis of Internet Content Delivery Systems Stefan Saroiu, Krishna P. Gommadi, Richard J. Dunn, Steven D. Gribble, and Henry M. Levy Proceedings of.
EECS Presentation Web Tap: Intelligent Intrusion Detection Kevin Borders.
ICNP'061 Benefit-based Data Caching in Ad Hoc Networks Bin Tang, Himanshu Gupta and Samir Das Computer Science Department Stony Brook University.
ICNP'061 Benefit-based Data Caching in Ad Hoc Networks Bin Tang, Himanshu Gupta and Samir Das Department of Computer Science Stony Brook University.
Kyushu University Graduate School of Information Science and Electrical Engineering Department of Advanced Information Technology Supervisor: Professor.
Implementing ISA Server Caching. Caching Overview ISA Server supports caching as a way to improve the speed of retrieving information from the Internet.
Adaptive Content Delivery for Scalable Web Servers Authors: Rahul Pradhan and Mark Claypool Presented by: David Finkel Computer Science Department Worcester.
Tcl Agent : A flexible and secure mobile-agent system Paper by Robert S. Gray Dartmouth College Presented by Vipul Sawhney University of Pennsylvania.
Peer-to-peer Multimedia Streaming and Caching Service by Won J. Jeon and Klara Nahrstedt University of Illinois at Urbana-Champaign, Urbana, USA.
How to Own the Internet in your spare time Ashish Gupta Network Security April 2004.
A Study on Mobile P2P Systems Hongyu Li. Outline  Introduction  Characteristics of P2P  Architecture  Mobile P2P Applications  Conclusion.
Introduction to client/server architecture
Middleware for P2P architecture Jikai Yin, Shuai Zhang, Ziwen Zhang.
Overview SAP Basis Functions. SAP Technical Overview Learning Objectives What the Basis system is How does SAP handle a transaction request Differentiating.
A Secure Protocol for Spontaneous Wireless Ad Hoc Networks Creation.
Windows 2000 Advanced Server and Clustering Prepared by: Tetsu Nagayama Russ Smith Dale Pena.
High Performance User-Level Sockets over Gigabit Ethernet Pavan Balaji Ohio State University Piyush Shivam Ohio State University.
Workload-driven Analysis of File Systems in Shared Multi-Tier Data-Centers over InfiniBand K. Vaidyanathan P. Balaji H. –W. Jin D.K. Panda Network-Based.
Distributed Systems: Concepts and Design Chapter 1 Pages
Architectures of distributed systems Fundamental Models
A Measurement Based Memory Performance Evaluation of High Throughput Servers Garba Isa Yau Department of Computer Engineering King Fahd University of Petroleum.
Kiew-Hong Chua a.k.a Francis Computer Network Presentation 12/5/00.
1 Evaluation of Cooperative Web Caching with Web Polygraph Ping Du and Jaspal Subhlok Department of Computer Science University of Houston presented at.
Distributed Databases
Distributed database system
Empirical Quantification of Opportunities for Content Adaptation in Web Servers Michael Gopshtein and Dror Feitelson School of Engineering and Computer.
A P2P-Based Architecture for Secure Software Delivery Using Volunteer Assistance Purvi Shah, Jehan-François Pâris, Jeffrey Morgan and John Schettino IEEE.
PwC New Technologies New Risks. PricewaterhouseCoopers Technology and Security Evolution Mainframe Technology –Single host –Limited Trusted users Security.
Jini Architecture Introduction System Overview An Example.
Kemal Baykal Rasim Ismayilov
Jini Architectural Overview Li Ping
Search Engine using Web Mining COMS E Web Enhanced Information Mgmt Prof. Gail Kaiser Presented By: Rupal Shah (UNI: rrs2146)
1 My Dream of Jini Fabio Kon Jalal Al-Muhtadi Roy Campbell M. Dennis Mickunas Department of Computer Science University of Illinois at.
Web Server By Bhupendra Ratha, Lecturer School of Library and Information Science Devi Ahilya University, Indore
Peer-to-Peer Systems: An Overview Hongyu Li. Outline  Introduction  Characteristics of P2P  Algorithms  P2P Applications  Conclusion.
WebWatcher A Lightweight Tool for Analyzing Web Server Logs Hervé DEBAR IBM Zurich Research Laboratory Global Security Analysis Laboratory
Mobile Analyzer A Distributed Computing Platform Juho Karppinen Helsinki Institute of Physics Technology Program May 23th, 2002 Mobile.
1 Evaluation of Cooperative Web Caching with Web Polygraph Ping Du and Jaspal Subhlok Department of Computer Science University of Houston presented at.
Lecture # 10 Processors Microcomputer Processors.
Presented by Deepak Varghese Reg No: Introduction Application S/W for server load balancing Many client requests make server congestion Distribute.
August 2001Copyright UserLand Software 2001 Next Generation Internet Software How Broadband will be used John Robb, President and COO UserLand Software.
Geethanjali College Of Engineering and Technology Cheeryal( V), Keesara ( M), Ranga Reddy District. I I Internal Guide Mrs.CH.V.Anupama Assistant Professor.
VIRTUAL NETWORK COMPUTING SUBMITTED BY:- Ankur Yadav Ashish Solanki Charu Swaroop Harsha Jain.
Cooperative Caching in Wireless P2P Networks: Design, Implementation And Evaluation.
NetFlow Analyzer Best Practices, Tips, Tricks. Agenda Professional vs Enterprise Edition System Requirements Storage Settings Performance Tuning Configure.
An example of peer-to-peer application
Building Distributed Educational Applications using P2P
Introduction to PHP FdSc Module 109 Server side scripting and
Authors: Sajjad Rizvi, Xi Li, Bernard Wong, Fiodar Kazhamiaka
Introduction to client/server architecture
University of Technology
#01 Client/Server Computing
CPSC 641: WAN Measurement Carey Williamson
Carey Williamson Department of Computer Science University of Calgary
CS 286 Computer Organization and Architecture
Presentation transcript:

Overhaul: Extending HTTP to Combat Flash Crowds Jay A. Patel & Indranil Gupta Distributed Protocols Research Group Department of Computer Science University of Illinois at Urbana-Champaign (UIUC) Urbana, Illinois, USA

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign2 Introduction  Flash crowd: A stampede of unexpected visitors  Occurs regularly due to linkage from popular news feeds, web logs, etc. Popularly termed “Slashdot effect”  Victim sites become unresponsive Perception of dysfunction

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign3 Example: MSNBC MSNBC home page December 14, 2003

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign4 Motivation  Problem Unpredictable, yet frequent Brief period of time Thousand-fold increase in traffic  Two naïve solutions Overly insure on resources Shut down web site

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign5 Current Solutions  Architectural Changes SEDA Capriccio ESI  Protocol Modifications DHTTP Web Booster  Cooperative Sharing Squirrel Kache Backslash BitTorrent

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign6 HTTP: Regular Interaction Client Server GET Request Response Document Header

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign7 Overhaul: Overview  Protocol change HTTP extension, no modification  5 new tags added, 1 slightly modified Backwards compatible  Key concept: chunking Characteristic of the web applied to individual documents m chunks per document  P2P distribution framework Voluntary Ad hoc, not DHT based Key benefit: parallel resource discovery

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign8 Overhaul: Design Client Server Client #1 #2 #4 #3 HTTP Request with Overhaul support tag Chunked Response with Overhaul headers Peers exchange chunks to fetch the complete document Ad hoc peer network

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign9 Details: Client/Server Interaction  Initial request by client Supports: Overhaul $port $speed  Response by server in Overhaul mode i th chunk transmitted in sequential order Signatures of other m-1 chunks for verification Initial Overhaul network membership list  n most-recent Overhaul clients  List maintained at server (updated with every request)

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign10 Details: Peer Clients’ Interaction  Clients contact other peer members To fetch remaining chunks To discover new peers  Aggregate membership list by swapping information  1-hop random walk discovery process  Resource discovery Lookup documents on a busy Overhaul server  Contact peers randomly on membership list  INFO $host.tld

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign11 Implementation  Server Apache/2.0 HTTP server Module: mod_overhaul  Client Java HTTP Proxy Cross platform Universal client support

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign12 Testing Methodology: Server  Server machine 2.5 GHz AMD Athlon XP+ 1 GB RAM  Client machine 650 MHz Pentium III 320 MB RAM  Same network equipment  25 concurrent fetches  ApacheBench utility

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign13 Results: Chunking (Fixed Size) Document: 10 KB Concurrency: 25 Regular HTTP 512-byte chunks 2048-byte chunks Overhaul mode requires the server to send only a single chunk

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign14 Results: Chunking (Maximum Count) Regular HTTP 6 chunks 12 chunks 24 chunks Document: 50 KB Concurrency: 25

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign15 Results: Overhaul vs. Regular Regular HTTP 6 chunks 12 chunks Concurrency: 25 Minimum chunk size: 512-bytes

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign16 Testing Methodology: Client  Cluster of workstations 25 homogenous PCs  2.8 GHz Intel Pentium 4  1 GB RAM Same network equipment  Two experiments Concurrent: single document Staggered: multiple documents

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign17 Results: Single Document  Large document: 50 KB (12 chunks)  Server condition: concurrent fetches + competition  Overhaul requests: concurrently only using 24 Overhaul-aware clients Regular requests Overhaul mode Fastest1 sec6 secs Slowest32 secs9 secs Average9 secs7 secs Server bandwidth usage in Overhaul mode: 1/12 th of regular requests

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign18 Results: Multiple Documents  8 documents: 110 KB total (12 chunks)  Server condition: concurrent fetches + competition  Overhaul requests staggered 1 st stage: 12 concurrent fetches, fetch all documents 2 nd stage: 12 concurrent fetches, fetch index document only Regular requests Overhaul mode Fastest1 sec14 secs Slowest∞28 secs Average23 secs * 18 secs Server bandwidth usage in Overhaul mode : 1/18 th of regular requests * indicates completed requests

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign19 Limitations  Both client and server must be Overhaul aware  Requires critical mass to be maintained to remain effective n clients > m chunks  More responsibilities for the client  Possible security implications

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign20 Conclusion  Saves resources Bandwidth  The bigger the crowd, the lower the per capita usage Response time  Faster turnaround for both server and client  Getting wide spread acceptance Marginal cost Protocol extension requires industry and standards push

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign21 Overhaul Vs. BitTorrent  Specifically intended for flash crowds Feasible for short durations Small document size  Tightly integrated for HTTP Another server/software not required Resource discovery: built-in notion of related documents

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign22 Regarding Greedy Clients  Voluntary network  Must increase membership list to fetch document(s) faster Forces communication and sharing  Future work Trust score matrix based on sharing

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign23 Heterogeneous Networks  Problem Connections are heterogeneous  Solutions Clustering of clients Super nodes Client

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign24 Document Selection  Only a partial set of documents are affected by a flash crowd  Must implement selective Overhaul mode  Automatic selection Active monitoring Server Large collection of documents reside on the server Documents fetched by a flash crowd

Distributed Protocols Research Group, Department of Computer Science, University of Illinois at Urbana-Champaign25 Dynamic Documents  Flash crowds especially frequent during big events and news  Characteristic: rapidly, changing data  Solutions Time stamping Expiration of chunks Inter-network refresh from peers