Container cloud cluster resource utilization optimization method based on deep reinforcement learning

Abstract

The invention provides a container cloud cluster resource utilization optimization method based on deep reinforcement learning. The method comprises the steps of preprocessing original load data and assembling the original load data into an input state s; constructing a deep Q network model, inputting the input state s into the deep Q network model, randomly selecting an action a by the deep Q network model according to a certain probability, or selecting an action a enabling the deep Q network model to be optimal, and executing overselling ratio prediction once; evaluating the selected actiona through a reward function to obtain a reward r and entering a next state s'; forming a quadruple by the input state s, the action a, the reward r and the next state s', and putting the quadruple into a cache as a training sample; when a preset training interval is reached, sampling e training samples from the cache and inputting the samples into the deep Q network model for training, and updating parameters of the deep Q network model; and after the deep Q network model is subjected to E rounds of training, applying the deep Q network model of which the parameters are updated to determine an overselling strategy.

Description

technical field [0001] The present invention relates to the technical field of cloud computing resource management, and more specifically, to a container cloud cluster resource utilization optimization method based on deep reinforcement learning. Background technique [0002] Docker is an open source application container engine. With the wide application of docker container technology in application development, testing and release, Google first proposed kubernetes (K8s) in 2015, a distributed architecture solution based on docker container technology. K8s provides complete cluster management capabilities, such as multi-level security protection and access control, transparent service registration and service discovery mechanisms, and multi-granularity resource management capabilities. In addition, k8s also provides a built-in load balancer and scalable resource automatic scheduling capability. The scheduling capability is provided by the built-in scheduler and requires the...

AI Technical Summary

Problems solved by technology

However, the advantage of this static scheduling method is that it is simple and effective enough, but there is a problem that it is easy to lead to low resource utilization of the cluster, which is also a common problem in the industry

Images