Self Healing System with Kubernetes and AI

Description

This project demonstrates a self-healing Kubernetes system driven by AI-based anomaly detection. It collects metrics from Prometheus, exposes a custom metric (anomaly_count), and uses Prometheus Adapter plus Horizontal Pod Autoscaler to scale pods automatically when anomalies spike. This setup showcases a dynamic, AI-powered approach to resilience and auto-scaling in Kubernetes.

Languages and Utilities Used

• Python
• Docker
• Kubernetes
• Helm
• Prometheus & Prometheus Adapter
• Horizontal Pod Autoscaler (HPA)

Project Walk-through

AI Model & Script


Containerized for Easy Deployment


Kubernetes Deployment & Service


Scraping Custom Metric & Monitoring with Prometheus


Self-Healing System with HPA