Realizing fault resilience in Web-server cluster
Proceedings of the 2000 ACM/IEEE conference on Supercomputing
The state of the art in locally distributed Web-server systems
ACM Computing Surveys (CSUR)
Dynamic Load Balancing on Web-Server Systems
IEEE Internet Computing
IEEE Transactions on Knowledge and Data Engineering
ICPP '00 Proceedings of the Proceedings of the 2000 International Conference on Parallel Processing
Cyclone: A High-Performance Cluster-Based Web Server with Socket Cloning
Cluster Computing
Scalable load balancing on distributed web servers using mobile agents
Journal of Parallel and Distributed Computing - Scalable web services and architecture
Dynamic traffic controls for web-server networks
Computer Networks: The International Journal of Computer and Telecommunications Networking
Efficient support for content-based routing in web server clusters
USITS'99 Proceedings of the 2nd conference on USENIX Symposium on Internet Technologies and Systems - Volume 2
Enabling fault resilience for web services
Computer Communications
Design, implementation, and performance of a load balancer for SIP server clusters
IEEE/ACM Transactions on Networking (TON)
| Hi-index | 0.00 |
Providing World Wide Web or FTP service on a small scale is a well-solved problem, but scaling server technology for a site that accepts millions of connections per day can easily push multiple machines and networks to the bleeding edge. In this case study, we describe concrete configuration techniques that helped us get the best possible performance out of server resources at Netscape Communications. We designed, implemented, and babysat the Netscape Web site from its inception until traffic grew to about five million hits per day. The site now gets more than 120 million hits per day. We describe many iterations through the loop of watching the load increase, finding various failures, and fixing what was broken. We hope our experience will help readers configure a high-performance server from the start and then know what to look for when it becomes overloaded