Module 07: Docker and Kubernetes

Goal

Package AI systems consistently and deploy them to a local Kubernetes environment with health checks, scaling, configuration, and secrets.

Docker Topics

• Dockerfile.
• Images.
• Containers.
• Volumes.
• Networks.
• Multi-stage builds.
• Image security.
• Registries.

Kubernetes Topics

• Pod.
• Deployment.
• Service.
• Ingress.
• ConfigMap.
• Secret.
• Namespace.
• Horizontal Pod Autoscaler.
• Persistent Volume.
• Resource limits.
• Readiness and liveness probes.

Study Tasks

• Dockerize a FastAPI or Node API.
• Add a multi-stage Dockerfile.
• Add Docker Compose for local services.
• Deploy backend, database, cache, and vector database locally.
• Create Kubernetes manifests.
• Add health checks.
• Add resource requests and limits.
• Add horizontal scaling.

Project

Containerize projects 04-enterprise-rag-platform and 05-erp-ai-agent, then use the deployment patterns in projects/07-production-cloud-ai-platform/.

Tips

• Keep images small and reproducible.
• Never bake secrets into images.
• Use health endpoints before autoscaling.
• Test graceful shutdown.

Completion Checklist

• Can write a production-style Dockerfile.
• Can run a multi-service stack locally.
• Can deploy to a local Kubernetes cluster.
• Can configure secrets safely.
• Can explain resource limits and autoscaling.