Microservice Deployment¶

🧠 Objectives¶

By the end of this session, students will:

• Understand monolithic vs. microservices architecture.
• Learn how containers enable microservices.
• Explore deployment strategies like rolling updates, blue-green, canary.

📖 Monolith vs. Microservices¶

• Monolithic app = One big application, tightly coupled.
• Microservices = Multiple smaller services, loosely coupled, each responsible for one function.

Example:

• Monolith = single e-commerce app.
• Microservices = independent services (cart, payments, search, user auth).

🐳 Why Containers for Microservices?¶

• Encapsulate dependencies.
• Ensure consistency across environments.
• Easy scaling (scale only the service you need).

🚀 Deployment Strategies¶

• Rolling Update: Replace old pods with new gradually.
• Blue-Green Deployment: Run two environments (blue = current, green = new). Switch traffic once tested.
• Canary Release: Release to a small % of users first, then expand.

⚠️ Challenges in Microservices¶

• Service discovery: How do services find each other?
• Networking: Secure and efficient communication.
• Monitoring & Logging: Observability across multiple services.
• Data consistency: Distributed databases.

🔧 Tools for Microservices¶

• Kubernetes → Orchestration platform.
• Istio/Linkerd (Service Meshes) → Advanced routing, observability, security.
• Prometheus + Grafana → Monitoring.
• Elastic Stack (ELK) → Logging and search.

📊 Microservices Deployment Flow¶

flowchart LR
    A[Developer builds microservice] --> B[Container Image Built]
    B --> C[Push to Docker Hub]
    C --> D[Kubernetes Cluster]
    D --> E[Pods Running Microservice]
    E --> F[Service Mesh/Ingress]
    F --> G[End User Access]

🌍 Real-World Example¶

• Netflix:

  • Runs thousands of microservices for streaming, recommendations, payments.
  • Uses Kubernetes + custom orchestration.

• Amazon:

  • Each service (cart, payments, suggestions) runs as microservices.
  • Independent teams manage each service.