Kubernetes container cluster node selection method based on hybrid element heuristic algorithm

Abstract

The invention discloses a Kubernetes container cluster node selection method based on a hybrid element heuristic algorithm, wherein the method comprises the steps: building a cluster node cost model in a cloud environment, enabling the model to enable the total cost of all nodes of a cluster to be minimum, solving an optimal node selection scheme through the hybrid element heuristic algorithm based on simulated annealing and particle swarm optimization. The node selection scheme solved by the model is subjected to actual feasibility verification by using a scheduling strategy based on Kubernetes, so that the node selection with the optimal cluster cost of the Kubernetes container cluster on the premise of meeting the working requirements is realized. According to the method, the use cost of the cloud server of the cluster can be reduced through an optimized cluster node selection scheme on the premise of meeting the deployment requirement of the cluster container according to the server model and selling price provided by a cloud manufacturer and the container resource requirement needing to be deployed in a future period of time of the Kubernetes container cluster.

Description

technical field [0001] The present invention relates to a method for selecting Kubernetes container cluster nodes based on a hybrid meta-heuristic algorithm, more specifically, using a particle swarm algorithm and a hybrid meta-heuristic algorithm based on simulated annealing and particle swarm optimization algorithms to meet the premise of cluster container deployment Next, select the cluster node that minimizes the cloud server cost that the cluster needs to pay. Background technique [0002] Container technology first appeared in 2001. Of course, at this time, only a Namespace module was added to the Linux kernel, and a complete container model had not yet appeared. In 2004, Google began to use Borg to manage and maintain distributed clusters scattered around the world. It is the predecessor of Docker; in October 2007, Google submitted a milestone Cgroups module to the Linux kernel, and in 2008 based on the above, it promoted the official release of LXC (Linux Container);...

AI Technical Summary

Problems solved by technology

However, many cloud platform container service users find it difficult to determine how to select nodes when selecting cluster nodes, so as to minimize the cost of cloud servers that should be paid by the cluster on the premise of meeting the resource requirements of container deployment.

At present, the common practice is to use the cluster elastic scaling function provided by the cloud platform to automatically expand or reduce the use of nodes for the cluster. However, automatic expansion often cannot be based on the resource gap of the actual container deployment, the type of cloud server resources, and The price is comprehensively considered to select nodes, which will have the defect that the selection of expanded nodes is not the most economical and suitable

Images