Handling Kubernetes clusters is a troublesome task, especially for beginner DevOps engineers. Most of the time, the tools in use require a command-line knowledge and give the information in a complicated manner, thereby making the process of learning and troubleshooting very slow. However, the current project presents an AI-driven DevOps Assistant which, through a web-based dashboard, brings about an easy and pleasant Kubernetes interaction and controlling thus monitoring and management. The system is an easy-to-use interface where users are able to monitor workloads, logs, and metrics in a visual way without any need for the command line. The users can also perform cluster operations via structured queries or simply by asking questions to an AI assistant. The front end is developed in React while the backend is developed in Node.js to establish communication with the Kubernetes cluster via APIs. Metrics for CPU and memory are collected utilizing Prometheus and visualized through Grafana charts. A database-driven system manages both user authentication and role-based access. The project is aimed at bringing in Kubernetes the simplicity of understanding, and management in controlled DevOps environments. The scope of the project must not be limited just to the mentioned advanced AI functionalities, anomaly detection, and massive-scale production system support in the future.
Tools: Kubernetes, Docker, GitHub, Git, React.js, Node.js, FastAPI, REST APIs
Department: Department of Computer Science
Poster
