Application-Layer Anycasting By Samarat Bhattacharjee et al. Presented by Matt Miller September 30, 2002.

Slides:



Advertisements
Similar presentations
1 Data-Oriented Network Architecture (DONA) Scott Shenker (M. Chowla, T. Koponen, K. Lakshminarayanan, A. Ramachandran, A. Tavakoli, I. Stoica)
Advertisements

Push Technology Humie Leung Annabelle Huo. Introduction Push technology is a set of technologies used to send information to a client without the client.
Internetworking II: MPLS, Security, and Traffic Engineering
Cs/ee 143 Communication Networks Chapter 6 Internetworking Text: Walrand & Parekh, 2010 Steven Low CMS, EE, Caltech.
Consistency and Replication Chapter 7 Part II Replica Management & Consistency Protocols.
1 Content Delivery Networks iBAND2 May 24, 1999 Dave Farber CTO Sandpiper Networks, Inc.
1 Routing and Scheduling in Web Server Clusters. 2 Reference The State of the Art in Locally Distributed Web-server Systems Valeria Cardellini, Emiliano.
Small-world Overlay P2P Network
An Engineering Approach to Computer Networking
Internet Networking Spring 2006 Tutorial 12 Web Caching Protocols ICP, CARP.
Application Layer Anycasting: A Server Selection Architecture and Use in a Replicated Web Service Presented in by Jayanthkumar Kannan On 11/26/03.
1 Spring Semester 2007, Dept. of Computer Science, Technion Internet Networking recitation #13 Web Caching Protocols ICP, CARP.
Kyushu University Graduate School of Information Science and Electrical Engineering Department of Advanced Information Technology Supervisor: Professor.
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.
1 Routing as a Service Karthik Lakshminarayanan (with Ion Stoica and Scott Shenker) Sahara/i3 retreat, January 2004.
70-293: MCSE Guide to Planning a Microsoft Windows Server 2003 Network, Enhanced Chapter 7: Planning a DNS Strategy.
1 Web Content Delivery Reading: Section and COS 461: Computer Networks Spring 2007 (MW 1:30-2:50 in Friend 004) Ioannis Avramopoulos Instructor:
DOMAIN NAME SYSTEM. Introduction  There are several applications that follow client server paradigm.  The client/server programs can be divided into.
DNS. Outline r Domain Name System r DNS Hierarchy r Resolution.
(part 3).  Switches, also known as switching hubs, have become an increasingly important part of our networking today, because when working with hubs,
FIREWALL TECHNOLOGIES Tahani al jehani. Firewall benefits  A firewall functions as a choke point – all traffic in and out must pass through this single.
11.1 © 2004 Pearson Education, Inc. Exam Managing and Maintaining a Microsoft® Windows® Server 2003 Environment Lesson 11: Introducing WINS, DNS,
1 Content Distribution Networks. 2 Replication Issues Request distribution: how to transparently distribute requests for content among replication servers.
1 Study on Adaptation of CDN Request-Routing to Scalable Conference System Toshiyuki KAWASAKI* Koji OKAMURA** * Graduate School of Information Science.
Active Network Applications Tom Anderson University of Washington.
1 Chapter 27 Internetwork Routing (Static and automatic routing; route propagation; BGP, RIP, OSPF; multicast routing)
Chapter 16 – DNS. DNS Domain Name Service This service allows client machines to resolve computer names (domain names) to IP addresses DNS works at the.
Inter-domain AMT Multicast Use Case Discussion Proposal for AMT Multicast Source-AMT Connectivity Model For Inter-connected Networks (AS’s) 1.
{ Content Distribution Networks ECE544 Dhananjay Makwana Principal Software Engineer, Semandex Networks 5/2/14ECE544.
23-Support Protocols and Technologies Dr. John P. Abraham Professor UTPA.
Unwanted Link Layer Traffic in Large IEEE Wireless Network By Naga V K Akkineni.
Professor OKAMURA Laboratory. Othman Othman M.M. 1.
Computer Networks. IP Addresses Before we communicate with a computer on the network we have to be able to identify it. Every computer on a network must.
DNS (Domain Name System) Protocol On the Internet, the DNS associates various sorts of information with domain names. A domain name is a meaningful and.
Crawlers and Spiders The Web Web crawler Indexer Search User Indexes Query Engine 1.
1 Application Layer Lecture 6 Imran Ahmed University of Management & Technology.
1 Chapter 6: Proxy Server in Internet and Intranet Designs Designs That Include Proxy Server Essential Proxy Server Design Concepts Data Protection in.
1 Routing. 2 Routing is the act of deciding how each individual datagram finds its way through the multiple different paths to its destination. Routing.
Submission doc.: IEEE /1015r1 September 2015 Guido R. Hiertz et al., EricssonSlide 1 Proxy ARP in ax Date: Authors:
Professor OKAMURA Laboratory. Othman Othman M.M. 1.
Module 12: Routing Fundamentals. Routing Overview Configuring Routing and Remote Access as a Router Quality of Service.
IP Forwarding.
 An Internet Protocol address (IP address) is a numerical label assigned to each device (e.g., computer, printer) participating in a computer network.
TOMA: A Viable Solution for Large- Scale Multicast Service Support Li Lao, Jun-Hong Cui, and Mario Gerla UCLA and University of Connecticut Networking.
© McLean HIGHER COMPUTER NETWORKING Lesson 4: Domain Name Service Description of domain names and name resolution Domain name servers and domain.
Networks and Protocols CE Week 7b. Routing an Overview.
Distributed Computing A Programmer’s Perspective.
Eric Tryon Brian Clark Christopher McKeowen. System Architecture The architecture can be broken down to three different basic layers Stub/skeleton layer.
DNS DNS overview DNS operation DNS zones. DNS Overview Name to IP address lookup service based on Domain Names Some DNS servers hold name and address.
Push Technology Humie Leung Annabelle Huo. Introduction Push technology is a set of technologies used to send information to a client without the client.
Ασύρματες και Κινητές Επικοινωνίες Ενότητα # 10: Mobile Network Layer: Mobile IP Διδάσκων: Βασίλειος Σύρης Τμήμα: Πληροφορικής.
70-293: MCSE Guide to Planning a Microsoft Windows Server 2003 Network, Enhanced Chapter 6: Planning, Configuring, And Troubleshooting WINS.
CIS679: Anycast r Review of Last lecture r Network-layer Anycast m Single-path routing for anycast messages r Application-layer anycast.
John S. Otto Mario A. Sánchez John P. Rula Fabián E. Bustamante Northwestern, EECS.
Chapter 4: server services. The Complete Guide to Linux System Administration2 Objectives Configure network interfaces using command- line and graphical.
A Fragmented Approach by Tim Micheletto. It is a way of having multiple cache servers handling data to perform a sort of load balancing It is also referred.
Network Topologies for Scalable Multi-User Virtual Environments Lingrui Liang.
Domain Name System: DNS To identify an entity, TCP/IP protocols use the IP address, which uniquely identifies the Connection of a host to the Internet.
Naming for Mobile Systems
Scaling Network Load Balancing Clusters
IMPLEMENTING NAME RESOLUTION USING DNS
Mohammad Malli Chadi Barakat, Walid Dabbous Alcatel meeting
Internet Networking recitation #12
Net 323 D: Networks Protocols
Managing Online Services
Net 323 D: Networks Protocols
Outline Review of Quiz #1 Distributed File Systems 4/20/2019 COP5611.
EE 122: Lecture 22 (Overlay Networks)
Presentation transcript:

Application-Layer Anycasting By Samarat Bhattacharjee et al. Presented by Matt Miller September 30, 2002

Motivation Given that multiple replicas of a service are available, how do we connect to the “best” one for a particular client? Anycast has been defined as a service and a framework specified for the IP layer. How can we specify an anycast framework at the application layer?

Key Contributions Presents arguments why anycast should not be implemented at the network layer Provides an application layer framework for implementing anycast Enumerates possible filters and metrics that could be used and how they could be supported Adapts server pushing for updating state information that trades off accuracy for control overhead

Limitations of Network Layer Anycast Address space issues in IPv4 – Use existing addresses and make identification difficult – Use a separate set of addresses and risk inefficient routing Requires router support to avoid delivering to multiple hosts

Limitations of Network Layer Anycast Most protocols would like all data for a connection delivered to one IP address once a service is found “Best” only refers to shortest hop count. At the application layer, many other metrics (possibly user-defined) may be applied.

Service Location How to find a service – Multicast to find it – Use name server architectures – Caching a resource location where it is frequently accessed How to find the “best” service – Gather information from servers and efficiently search through it – Servers periodically push their local state

Replicated Services Replicated services are equivalent in content and/or functionality from an application perspective Compute servers are machines which are capable of running a particular computation – Server statistics such as CPU load may be an important criteria

Anycast Domain Names Anycast Domain Names (ADN) identify an anycast group of potentially dynamic IP addresses The group could also be specified as domain names or aliases instead of IP addresses

Anycast Name Resolution

Anycast Name Resolution (2) Works like DNS server A service and domain name are specified The domain name is resolved by hierarchtically querying ADN servers until an authoritative response or cached entry is found The ADN maintains a list of IP addresses for a service and associated metrics

Anycast Name Resolution (3)

Filtering The local ADN resolver can filter addresses given by authoritative entity The client must handle multiple or no addresses being returned by the resolver Three proposed filters – Content-independent – Metric-based – Policy-based

Content-Independent Filter Random selection of a member Return all members of the group Return a subset of n members of the group

Metric-Based Filter Select the best member according to a single metric Select the best member according to a function of multiple metrics Select the best by sequentially applying filters

Policy-Based Filters Vague description, not based on performance measurements Generally, it would be a boolean function which determines whether an address meets a policy criteria or not

Filter Issues How can clients tell resolvers what filter to run – Use well-known identifiers – Allow clients to give procedural description How is it implemented – Create a new function with filters – Specialized domain names (Metric-Qualified ADN) Backwards compatible E.g. ServerLoad.wwwnews%cc.gatech.edu.any

Metric-Qualified ADN Implementation

Metrics Metrics are relative rather than absolute Goal is to get reasonable accuracy without excessive network or server load Possible metrics – Latency – Throughput – Server Load

Metric Collection Remote Server Performance Probing – Proxies periodically query replicated servers to determine how potential clients would perform Server Push – Servers send data when changes occur – Could be multicast to all interested anycast resolvers

Metric Collection (2) Probe Locally-Maintained Server Performance – Probe request reads static data from the server which is periodically updated User Experience – Users give their preference of servers that have performed well in the past – No burden on server, but could be very inaccurate – Accuracy may be increased if clients share experiences

Metric Collection (3) Example of server push – If a particular metric has changed by more than a certain threshold in a time interval, push the data. – Otherwise, decrement the threshold by a specified amount. When it reaches zero, push the data. – Demonstrates the tradeoff in accuracy and control overhead

Metric Collection (4)

Conclusions Shows application-layer anycast is feasible and provides basic framework Gives clients more control in selecting servers and is easily extendible Opens issues – How to specify policy filters – How to provide client-to-server metrics in a scalable way – Stability in service location