Toward Smarter Hadoop’s Slaves Nodes by Deploying Game Theory Strategies

As each one of us recognized the high speed of evaluation in the technologies field, the result of which is generating a huge amount of data which is known as Big Data. For this Big Data, there is need for an efficient distributed system to process this data and here came the Hadoop. Hadoop is one of the important frameworks in distributed system that has many other applications to process data. There are many researchers working to improve the performance of Hadoop, but in all literature the main focus is to improve the Master node, while it is important to try and provide a smarter Slaves node that can cooperate or predict what other Slaves node strategies to pick up the Job. In this paper, our study focuses on Game theory domain such as Nash equilibrium and bargaining strategy, second is Artificial intelligent to propose a smarter system even slave nodes in it can take a decision, especially that the core of Hadoop System is adopting a pull-scheduling strategy. We believe our work is going to improve resources utilization and minimize the processing time of Jobs.