Dynamic load balancing algorithm for scalable heterogeneous web server cluster with content awareness
Ajay Tiwari, Priyesh Kanungo · 2010
The state-of-art web applications communicate and coordinate with number of geographically dispersed information resources providing information to huge number of users. Homogeneous server clusters are not capable of satisfying the growing demand of such applications including real time audio and video, PHP, JSP, ASP etc. Heterogeneity involves handling of low level interoperability issues e.g. mismatch of hardware, operating platforms, programming languages, database schema, topology etc. Scalable server cluster allows addition of new servers as the load increases without disrupting the services. Moreover, it also provides better reliability by gracefully transferring the load from server which is unavailable due to failure or for preventive maintenance. Heterogeneity with scalability makes the system more complex. The existing dynamic load balancing (DLB) algorithms are not directly applicable for distributed scheduling in such environments. In this paper, we propose a DLB algorithm for scalable heterogeneous server cluster using content awareness. The algorithm considers server's processing capability, queue length, utilization ratio etc. as load indices. As the cluster supports multiple services, at the primary level, we have used content awareness forwarding algorithm and at the secondary level, waited round robin algorithm has been used.