A Improved Task Scheduling Algorithm Baseon Weight Round-Robin Algorithm on Hadoop Platform
XU Shu-q · Computer and Information Technology · 2013
To increase job scheduling efficiency and system throughput in Hadoop platform,a job scheduling algorithm is proposed.The algorithm is based on weighted round robin(WRR) algorithm,and adds factors to improve the data locality of map tasks according to the characteristic of MapReduce.The methods reduce task execution time by improving the data locality of map tasks while maintaining fairness to some extent in every cycle.The experimental results show that compared with WRR algorithm,the proposed algorithm increases the scheduling ratio of local tasks,thus reduces the execution time of total jobs in the cloud computing system.