Multi Distributed Server Load Balancing for SaaS: With Parameterized CPU and Memory Statistics

The term load balancing refers to distribution of load among the various nodes in order to improve the job response time, resource utilization and avoid the scenario where some node doing more work and some are lightly loaded. The load balancing facilitates every node to perform the equal amount of work at a time. The proposed work shows the load balancing, where the clients request is distributed among the servers located at different geographical locations. Here the load balancing problem is solved by considering the statistics such as CPU, memory and location details. In this system initially, the user request will be sent to the server having the less CPU and memory usage, if the servers are equally loaded then the request is sent to nearest node with respect to user location. Thus load is moved from the heavily loaded node to lightly loaded node.