A scheduling algorithm based on Clara clustering
Ling Kuang, Lichen Zhang · AIP conference proceedings · 2017
Task scheduling is a key issue in cloud computing. A new algorithm for queuing task scheduling based on Clara clustering and SJF cloud computing is proposed to introduce the Clara clustering for the shortcomings of SJF algorithm load imbalance. The Clara clustering method prepares the task clustering based on the task execution time and the waiting time of the task, and then divides the task into three groups according to the reference point obtained by the clustering. Based on the number of tasks per group in the proportion of the total number of tasks assigned to the implementation of the quota. Each queue team will perform task scheduling based on these quotas and SJF. The simulation results show that the algorithm has good load balancing and system performance.