Task Scheduling Strategy For Computationally Intensive Batch Processing

In computationally intensive batch processing tasks,there is a task that runs for a long time and consumes a lot of CPU and memory resources,such as gene sequencing task.For this type of task scheduling strategy,task completion time is an important indicator to measure the quality of the scheduling strategy.When the task size and quantity are fixed,the task completion time is smaller,the system service quality is higher.A scheduling system for computationally intensive batch task was designed.The system utilizes an Improved Genetic Simulated Annealing algorithm(IGSA)for the task scheduling The algorithm combines the advantages of multi-point optimization of Genetic Algorithm(GA)and local search ability of Simulated Annealing(SA).The basic idea is to improve the of initialization the genetic algorithm,selection operator,crossover operator,mutation operator,crossover probability,mutation probability,and termination conditions of the algorithm.The initialization of the algorithm adopts a method of selecting some of the best solutions among the multiple random solutions as a part of initial solution,and in selection operator the optimal individual directly retain the next generation.The crossover operator adopts a double elite retention strategy.Probability of crossover and mutation probability is adaptive,and the algorithm terminates when optimal solution of the population does not change for several generations.Then multiple solutions generated by the improved genetic algorithm are used as input to the simulated annealing algorithm,and multiple solutions are simulated.The best solution of the simulated annealing algorithm is the final solution of the algorithm.The results of experiments show that the task scheduling using this algorithm can shorten the task completion time compared to the RoundRobin(RR)algorithm,the improved genetic algorithm as well as the simulated annealing algorithm,and make the system more balanced.