The Container Deployment Strategy Based on Stable Matching

The number of containers is usually far greater than the number of virtual machines in data center. In order to avoid the energy waste caused by the frequent migration of container after initial deployment, an efficient container initial deployment strategy is particularly important. Therefore, this paper studies a many-to-one stable matching algorithm and proposes a container placement strategy MLSM. We start at an initial hosting solution for containers to reduce their migration times by using a stable matching algorithm. This algorithm uses the similarity methods as the matching preference strategy for containers and virtual machines. The virtual machine preference list is sorted by the resource utilization rate. The simulation results show that the algorithm can reduce the energy consumption by an average of 12.8% compared with using FirstFit algorithm.