Research on optimization strategy of container orchestration technology for cloud computing environment

Abstract

Abstract Containers are becoming the new standard of technology in the cloud era, and a huge number of services are shifting from the traditional virtual machine-based monolithic architecture to the container-based cloud-native architecture. This paper explores the optimization of container orchestration technology in the cloud computing environment, mainly focusing on the optimization study of Kubernetes, a cloud resource orchestration tool for Docker containers. The problems of this container cloud platform were analyzed, and it was found that the orchestration strategy adopted by Kubernetes is a static orchestration strategy, which can easily cause redundancy of container resources. This paper suggests an optimization strategy that first analyzes the PSUB container resource allocation strategy and then presents a new container orchestration optimization strategy based on the PSUB container resource allocation strategy. This paper proposes a PBFDR strategy that utilizes CPU, memory, network bandwidth, and disk IO in a more reasonable and balanced manner by comparing and analyzing the optimization strategies of different container orchestration technologies. Introducing the PBFDR strategy into the Kubernetes container orchestration framework and comparing and analyzing it with the mainstream container orchestration frameworks Rancher and Docker Swarm, the balanced point of resource utilization of KLLPL is around 23%, and the load balancing degree in different deployment sizes is around 13.7%. The evidence indicates that container orchestration technology that is based on PBFDR policy can effectively enhance the utilization of cluster resources.