Proxy cache which is an important device to guarantee the quality of Web service works with a poor efficiency when hot-spot appears. To solve this problem, load balancing method should be applied. In this paper, a load balancing model based on ant colony behavior is put forward. In the model, agent monitors the status of proxy server it has been installed on. Based on that status, if it judges that the server is busy by hot-spot, the agent asks for help to agents on other servers. Then other servers will form cooperation according to status of themselves to share the requests caused by hot-spot, which can balance the load of system. The main idea of the model is described and the algorithm for load balancing of the model is introduced. The results of related experiments show that the model has a good ability of load balancing and storage space saving.
[1]
Zhang Sen.
Novel network traffic forecasting algorithm based on grey model and Markov chain
,
2007
.
[2]
G. Barish,et al.
World Wide Web caching: trends and techniques
,
2000,
IEEE Commun. Mag..
[3]
Satosi Kato,et al.
Adaptive load balancing content address hashing routing for reverse proxy servers
,
2004,
2004 IEEE International Conference on Communications (IEEE Cat. No.04CH37577).
[4]
Philip S. Yu,et al.
Load balancing and hot spot relief for hash routing among a collection of proxy caches
,
1999,
Proceedings. 19th IEEE International Conference on Distributed Computing Systems (Cat. No.99CB37003).