Multi-objective resource allocation of bag-of-tasks in heterogeneous computing system

With the development of new computer technologies and cloud computing, more and more heterogeneous computing systems are becoming an important platform to process largescale applications in many fields. How to effectively utilize various resources in heterogeneous computing system to process a large amount of independent tasks (also called bag-of-tasks) is a challenging problem, especially when considering more than one objectives of tasks. In this paper, we investigate a bi-objective resource allocation problem for bag-of-tasks in heterogeneous computing system. Since system cost and makespan of workloads are the major issue that need to be considered by the data center, we establish a bi-objective optimization model which minimizes the cost of system and the makespan of the bag-of-tasks, and then propose a novel effective bi-objective genetic algorithm based on MOEA/D to efficiently solve the bi-objective model. Finally, numerical simulation experiments are conducted, and experimental results verify the effectiveness of the proposed model and algorithm. © 2020 Computer Society of the Republic of China. All rights reserved.