Container Composition Scheduling Strategy Based on Erd?s-Rényi Model
Containers are a new type of virtualization technology.Compared with traditional virtual machines,containers have the advantages of low cost,high resource utilization,and easy migration.However,when scheduling existing container clusters,it is easy to ignore the internal correlation among containers,and simply treat them as static individuals,which will lead to an imbalance in cluster performance and an increase in energy consumption.To solve this problem,this paper proposes a container scheduling strategy based on the Erdős-Rényi model.Firstly,the correlation among containers is me-asured by introducing communication indicators;secondly,based on the improved Erdős-Rényi model,containers are grou-ped according to the degree of correlation;finally,according to the grouping results,a Pod group is considered as a whole for scheduling to maximize the retention.The correlation among containers makes the overall dynamic performance of the cluster more significant.Simulation results show that compared with the Spread algorithm,the scheduling strategy proposed in this paper reduces the task completion time by about 22％,and reduces energy consumption by about 5％to 30％.Therefore,the strategy model proposed in this paper can effectively help enterprises improve production efficiency and reduce costs.
NETL
NSTL
万方数据
