Experience with some principles for building an internet-scale reliable system

Authors:
Mike Afergan;Joel Wein;Amy LaMeyer
Affiliations:
Akamai and MIT;Akamai and Polytechnic University;Akamai
Venue:
WORLDS'05 Proceedings of the 2nd conference on Real, Large Distributed Systems - Volume 2
Year:
2005

Citing 4
Cited 7

Resilient overlay networks

SOSP '01 Proceedings of the eighteenth ACM symposium on Operating systems principles
Globally Distributed Content Delivery

IEEE Internet Computing
Recovery Oriented Computing: A New Research Agenda for a New Century

HPCA '02 Proceedings of the 8th International Symposium on High-Performance Computer Architecture
Toward recovery-oriented computing

VLDB '02 Proceedings of the 28th international conference on Very Large Data Bases

A scheme for solving Anycast scalability in IPv6

International Journal of Network Management
Improving performance on the internet

Communications of the ACM - Inspiring Women in Computing
Improving Performance on the Internet

Queue - Scalable Web Services
Implementing and operating an internet scale distributed application using service oriented architecture principles and cloud computing infrastructure

Proceedings of the 10th International Conference on Information Integration and Web-based Applications & Services
A communication model on implementing anycast service in mobile IPv6 networks

International Journal of Systems, Control and Communications
The Akamai network: a platform for high-performance internet applications

ACM SIGOPS Operating Systems Review
Scaling a monitoring infrastructure for the Akamai network

ACM SIGOPS Operating Systems Review

Quantified Score

Hi-index	0.00

Visualization

Abstract

We discuss the design methodology used to achieve commercial-quality reliability in the Akamai content delivery network. The network consists of 15,000+ servers in 1,100+ networks and spans 65+ countries. Despite the scale of the Akamai CDN and the unpredictable nature of the underlying Internet, we seek to build a system of extremely high reliability. We present some simple principles we use to assure high reliability, and illustrate their application. As there is some similarity-in-spirit between our implementation and recent trends in the research literature, we hope that sharing our experiences will be of value to a broad community.