A Container-Based of Edge Device Monitoring on Kubernetes

The concept of the Internet of Things has widely used in various applications in life from home to big cities that have a variety of edge devices to facilitate edge computing. This paper attempts to use Kubernetes to establish an environment for collecting the internal resource usage status of various node devices and implementing monitoring and visualization. The deep learning models are deployed on heterogeneous devices for performance evaluation and verification. In this experiment, Kubernetes used to maintain a Docker container that used to deploy all applications mentioned above. All the data then visualize through Grafana.