Preserving QoS of e-commerce sites through self-tuning: a performance model approach
Proceedings of the 3rd ACM conference on Electronic Commerce
An introduction to variable and feature selection
The Journal of Machine Learning Research
A smart hill-climbing algorithm for application server configuration
Proceedings of the 13th international conference on World Wide Web
Automated Cluster-Based Web Service Performance Tuning
HPDC '04 Proceedings of the 13th IEEE International Symposium on High Performance Distributed Computing
An analytical model for multi-tier internet services and its applications
SIGMETRICS '05 Proceedings of the 2005 ACM SIGMETRICS international conference on Measurement and modeling of computer systems
DrC4.5: Improving C4.5 by means of prior knowledge
Proceedings of the 2005 ACM symposium on Applied computing
Data Mining: Concepts and Techniques
Data Mining: Concepts and Techniques
Optimizing system configurations quickly by guessing at the performance
Proceedings of the 2007 ACM SIGMETRICS international conference on Measurement and modeling of computer systems
Automatic performance tuning for J2EE application server systems
WISE'05 Proceedings of the 6th international conference on Web Information Systems Engineering
A control-based middleware framework for quality-of-service adaptations
IEEE Journal on Selected Areas in Communications
Hi-index | 0.00 |
Web has become the main platform for the interchange of information and the transaction of commerce. The performance of a Web system can be greatly improved by tuning its configuration parameters. However, there are dozens or even hundreds of tunable parameters in one Web system, and tuning can be the tough work even for the most experienced server administrators. Traditional Web tuning methods only focus on two or three specified parameters, and can not provide an effective solution to the tuning problem when the number of parameters is large. In this paper, we propose a feature selection algorithm based on Information Gain criterion to find the key parameters of a Web system. The algorithm can pick out the parameters that significantly affect Web system performance. Therefore, the tuning approach can be simplified dramatically. We have carried out extensive experiments with different Web systems. The results show that the algorithm is effective in searching the most important parameters under different conditions and reducing the time cost of next tuning steps.