Node Performance Load Balancing Algorithm For Hadoop Cluster

Hadoop is an implementation of MapReduce algorithm[1]. It operates on Distributed, parallel systems. During the execution of MapReduce applications default hash Partitioning algorithm leads to uneven distribution of load among the reducer tasks. Because of this total execution time gets affected. Hence the algorithm is modified in which it collects the local information, combines them and distributes the data in order to avoid skew. Node performance Load Balancing algorithm is used to reduce the imbalance of the load among the reducers. Due to this the execution time taken to complete the task is reduced.