A new task scheduling method for 2 level load balancing in homogeneous distributed system
Lipika Datta · 2016
A distributed system consists of several autonomous nodes. In a distributed system some of the nodes may be overloaded due to a large number of job arrivals while other nodes may remain idle without any processing. The performance of a distributed system depends crucially on dividing up work effectively among the computing nodes. So a way is needed to share load across all the computing nodes. In centralized load balancing schemes, the load balancing decision is taken by a central server. So this scheme is not scalable. In contrast, fully distributed schemes are scalable, but they use local information. A hierarchical dynamic load balancing model is proposed in this paper where an ordinary node does not need to have a global system wide knowledge about the states of other nodes in the system. The proposed model is semi distributed as each cluster is represented by a cluster master.