Hadoop Job Scheduling Based on Mixed Ant-Genetic Algorithm

Massive job scheduling problem is an important research area in big data research era. This paper proposed self-adaptive job scheduling mechanism based on Ant-Genetic Algorithm aiming at improving convergence speed and accuracy by mutation strategy based on Ant Algorithm and efficient refinement within Genetic Algorithm. The experimental results show that the proposed algorithm can find the most suitable nodes for current jobs and improve efficiency of job scheduling on Hadoop clusters effectively.