Application Layer Anycasting: A Server Selection Architecture and Use in a Replicated Web Service Presented in 294-4 by Jayanthkumar Kannan On 11/26/03.

Slides:



Advertisements
Similar presentations
Dynamic Replica Placement for Scalable Content Delivery Yan Chen, Randy H. Katz, John D. Kubiatowicz {yanchen, randy, EECS Department.
Advertisements

Internet Indirection Infrastructure (i3 ) Ion Stoica, Daniel Adkins, Shelley Zhuang, Scott Shenker, Sonesh Surana UC Berkeley SIGCOMM 2002 Presented by:
Push Technology Humie Leung Annabelle Huo. Introduction Push technology is a set of technologies used to send information to a client without the client.
Pastry Peter Druschel, Rice University Antony Rowstron, Microsoft Research UK Some slides are borrowed from the original presentation by the authors.
© 2007 Cisco Systems, Inc. All rights reserved.Cisco Public ITE PC v4.0 Chapter 1 1 Addressing the Network – IPv4 Network Fundamentals – Chapter 6.
Receiver-driven Layered Multicast S. McCanne, V. Jacobsen and M. Vetterli University of Calif, Berkeley and Lawrence Berkeley National Laboratory SIGCOMM.
1 Content Delivery Networks iBAND2 May 24, 1999 Dave Farber CTO Sandpiper Networks, Inc.
1 June 2015 Validating Inter-Domain SLAs with a Programmable Traffic Control System Elisa Boschi
Receiver-driven Layered Multicast S. McCanne, V. Jacobsen and M. Vetterli SIGCOMM 1996.
An Engineering Approach to Computer Networking
SCAN: A Dynamic, Scalable, and Efficient Content Distribution Network Yan Chen, Randy H. Katz, John D. Kubiatowicz {yanchen, randy,
Web Caching Schemes1 A Survey of Web Caching Schemes for the Internet Jia Wang.
Resource Management – a Solution for Providing QoS over IP Tudor Dumitraş, Frances Jen-Fung Ning and Humayun Latif.
Beneficial Caching in Mobile Ad Hoc Networks Bin Tang, Samir Das, Himanshu Gupta Computer Science Department Stony Brook University.
Internet Indirection Infrastructure Ion Stoica UC Berkeley.
Goal: To build a ubiquitous and robust storage infrastructure Requirement: Scalability, availability, performance, robustness Solution: Dynamic object.
Peer-to-Peer Based Multimedia Distribution Service Zhe Xiang, Qian Zhang, Wenwu Zhu, Zhensheng Zhang IEEE Transactions on Multimedia, Vol. 6, No. 2, April.
Scalable Application Layer Multicast Suman Banerjee Bobby Bhattacharjee Christopher Kommareddy ACM SIGCOMM Computer Communication Review, Proceedings of.
1 AINA 2006 Wien, April th 2006 DiVES: A DISTRIBUTED SUPPORT FOR NETWORKED VIRTUAL ENVIRONMENTS The IEEE 20th International Conference on Advanced.
OSMOSIS Final Presentation. Introduction Osmosis System Scalable, distributed system. Many-to-many publisher-subscriber real time sensor data streams,
CS 268: Project Suggestions Ion Stoica February 6, 2003.
Anycast Jennifer Rexford Advanced Computer Networks Tuesdays/Thursdays 1:30pm-2:50pm.
Application-Layer Anycasting: A Server Selection Architecture and Use in a Replicated Web Service IEEE/ACM Transactions on Networking Vol.8, No. 4, August.
1 Drafting Behind Akamai (Travelocity-Based Detouring) AoJan Su, David R. Choffnes, Aleksandar Kuzmanovic, and Fabian E. Bustamante Department of Electrical.
Measurement-Based Server Selection within the Application-Layer Anycasting Architecture Mostafa H. Ammar College of Computing Georgia Institute of Technology.
Multicast Communication
1 Routing as a Service Karthik Lakshminarayanan (with Ion Stoica and Scott Shenker) Sahara/i3 retreat, January 2004.
1 Web Content Delivery Reading: Section and COS 461: Computer Networks Spring 2007 (MW 1:30-2:50 in Friend 004) Ioannis Avramopoulos Instructor:
Internet Indirection Infrastructure (i3) Ion Stoica, Daniel Adkins, Shelley Zhuang, Scott Shenker, Sonesh Surana UC Berkeley SIGCOMM 2002.
1CS 6401 Peer-to-Peer Networks Outline Overview Gnutella Structured Overlays BitTorrent.
Receiver-driven Layered Multicast Paper by- Steven McCanne, Van Jacobson and Martin Vetterli – ACM SIGCOMM 1996 Presented By – Manoj Sivakumar.
© 2009 AT&T Intellectual Property. All rights reserved. Multimedia content growth: From IP networks to Medianets Cisco-IEEE ComSoc Webinar. Sept. 23, 2009.
1 Content Distribution Networks. 2 Replication Issues Request distribution: how to transparently distribute requests for content among replication servers.
Redirection and Load Balancing
Oasis: Anycast for Any Service Michael J. Freedman Karthik Lakshminarayanan David Mazières in NSDI 2006 Presented by: Sailesh Kumar.
Developing Analytical Framework to Measure Robustness of Peer-to-Peer Networks Niloy Ganguly.
Professor OKAMURA Laboratory. Othman Othman M.M. 1.
CS An Overlay Routing Scheme For Moving Large Files Su Zhang Kai Xu.
1 Chapter 6: Proxy Server in Internet and Intranet Designs Designs That Include Proxy Server Essential Proxy Server Design Concepts Data Protection in.
Application-Layer Anycasting By Samarat Bhattacharjee et al. Presented by Matt Miller September 30, 2002.
Overcast: Reliable Multicasting with an Overlay Network CS294 Paul Burstein 9/15/2003.
Security for the Optimized Link- State Routing Protocol for Wireless Ad Hoc Networks Stephen Asherson Computer Science MSc Student DNA Lab 1.
Jonathan Walpole CSE515 - Distributed Computing Systems 1 Teaching Assistant for CSE515 Rahul Dubey.
Professor OKAMURA Laboratory. Othman Othman M.M. 1.
Streaming over Subscription Overlay Networks Department of Computer Science Iowa State University.
Anycast DNS. WatITis | Strengthening Collaboration | December 8, 2009 | Anycast DNS Outline Current Anycast routing Anycast implemented Problems resolved.
Using DHCPv6 for DNS Configuration in Hosts draft-ietf-droms-dnsconfig-dhcpv6-00.txt Ralph Droms.
Application of Content Computing in Honeyfarm Introduction Overview of CDN (content delivery network) Overview of honeypot and honeyfarm New redirection.
Live Streaming over Subscription Overlay Networks CS587x Lecture Department of Computer Science Iowa State University.
A Routing Underlay for Overlay Networks Akihiro Nakao Larry Peterson Andy Bavier SIGCOMM’03 Reviewer: Jing lu.
Fast Handoff for Seamless wireless mesh Networks Yair Amir, Clauiu Danilov, Michael Hilsdale Mobisys’ Jeon, Seung-woo.
TOMA: A Viable Solution for Large- Scale Multicast Service Support Li Lao, Jun-Hong Cui, and Mario Gerla UCLA and University of Connecticut Networking.
Beyond Server Selection: Challenges in Multiple-Origin Content Distribution Mostafa H. Ammar College of Computing Georgia Institute of Technology Atlanta,
Othman Othman M.M., Koji Okamura Kyushu University 1.
The Replica Location Service The Globus Project™ And The DataGrid Project Copyright (c) 2002 University of Chicago and The University of Southern California.
Introduction to Mobile IPv6
NUS.SOC.CS5248 Ooi Wei Tsang Course Matters. NUS.SOC.CS5248 Ooi Wei Tsang Deadlines 11 Oct: Survey Paper Due 18 Oct: Paper Reviews Due.
Push Technology Humie Leung Annabelle Huo. Introduction Push technology is a set of technologies used to send information to a client without the client.
CS 6401 Overlay Networks Outline Overlay networks overview Routing overlays Resilient Overlay Networks Content Distribution Networks.
Click to edit Master title style Multi-Destination Routing and the Design of Peer-to-Peer Overlays Authors John Buford Panasonic Princeton Lab, USA. Alan.
Content Delivery Networks: Status and Trends Speaker: Shao-Fen Chou Advisor: Dr. Ho-Ting Wu 5/8/
Ch 2. Application Layer Myungchul Kim
Internet Indirection Infrastructure (i3)
CHAPTER 3 Architectures for Distributed Systems
Host Multicast: A Framework for Delivering Multicast to End Users
Early Measurements of a Cluster-based Architecture for P2P Systems
AWS Cloud Computing Masaki.
EE 122: Lecture 22 (Overlay Networks)
An Engineering Approach to Computer Networking
Multicasting Unicast.
Presentation transcript:

Application Layer Anycasting: A Server Selection Architecture and Use in a Replicated Web Service Presented in by Jayanthkumar Kannan On 11/26/03

Outline Motivation, Problem Definition Architecture Components Mechanisms API Experimental Results Conclusions

Motivation Anycast is a useful network primitive for accessing replicated content Service Model: Reach “best” node of a dynamic set of nodes Better suited to Application Layer deployment Metrics might be application-specific and routing might not be reactive enough. Routers need to be modified for anycast IPv4 address space needs to be reserved Stateless nature of IP: want to select server on flow-level, not per-packet

Goal An architecture for application-level anycast Client API Allow general metrics

Outline Motivation, Problem Definition Architecture Components Mechanisms API Experimental Results Conclusions

Architecture Main components Resolvers: Responsible for resolving anycast domain name to IP address Modified client: Domain name format modified Modified servers: Aid in monitoring and measurement

Anycast Resolvers Application-aware DNS-like resolvers Authoritative server for each domain Translate anycast domain name into IP address State required: List of IP addresses in each anycast group. Metrics associated with each IP address Authoritative resolver maintains definitive state for its groups, others cache it

Resolver operation Convert anycast domain name to IP address based on client-specified function of metrics Domain name format % points to authoritative resolver How do resolvers keep metrics up to date? Three kinds of metrics Server characteristics dependent (eg:load) Client-to-Server Path dependent (eg:total latency) Server and Path dependent

Metrics Maintanence Resolvers probe server periodically Works well if client is close to resolver Can measure all metrics Server Push Servers publish load data to multicast group subscribed to by resolvers Can measure server characteristics Resolvers probe for well known file at server User experience No additional traffic required

Client API Client uses gethostbyname(domain name) Domain name = FILTER % Filter: Given metrics for each server, which one is preferred? Content-Independent: Independent of metrics Metric-based: Relative/Absolute value of metrics Policy-based: Any general function

Client API (2) Client specifies filter and anycast domain name to resolver Does resolver have up to date information about group? Yes: Runs filter, and returns IP address Else: asks authoritative server, cache response, and return answer to client

Client API (3)

Observations O(S * G) state at each resolver G = # of groups, S = # of servers in each group Probe traffic could overwhelm server Scaling depends on number of resolvers/number of clients Resolvers replicate authoritative server’s state Generalize to multi-hop paths Essentially RON Can we use overlay routing protocol? Servers send periodic updates in say, DV. DV updated along overlay Unlikely to be reactive enough. Could potentially scale much better

Outline Motivation, Problem Definition Architecture Components Mechanisms API Experimental Results Conclusions

Case Study Web service replication Metric: Response Time = Path Latency + Server Processing Delay Metric Maintanence Server Push: Server publishes delay per request Agent Probe: Agents request well-known file Hybrid Techique: Agent measurements calibrated with server push data (adj factor = R/S) Note: Server Push more frequent than Agent Probing Found that hybrid technique tracks varying response time well

Refinements Hack to prevent oscillations Define ES = Equivalent set ES = set of servers with nearly same response times Form ES = ES + Minimum Response Time server Remove servers whose response time exceeds min by leave threshold Add servers whose response time within [min,min + join threshold] Hysteris used in selection criteriaES found using selection criteria Resolver returns randomly one element in ES Optimal Parameters dependent on several factors Oscillation definitely still a possibility

Experiment Configuration

Experimental Setup 4 anycast servers (1 in UCLA, 2 in GATECH, 1 in Washington U.) 2 Resolvers (UMCP and GATECH) 20 Clients (4 in UMCP, 16 in GATECH) Experiments Random vs Anycast choice Mixture of random choice, anycast choice Effect of Join Threshold

Random vs Anycast

Partial Deployment

Parameter Choice

Conclusions Architecture not very scalable Contributions Probing mechanisms to determine response time accurately with minimum overhead Possibly useful in P2P work