Summary of "Lec 15: Microservices"

Core definition and principles

Microservices architecture structures an application as a collection of independently deployable, loosely coupled services. Each service is small, focused, and organized around a single business capability (adhering to the Single Responsibility Principle).

Communication: services typically interact via REST APIs and are not tightly coupled to each other’s implementations.
Technology heterogeneity: different services can use different languages and frameworks (e.g., Java, Python).

Key characteristics and best practices

Small and focused services (one task or capability per service).
Decentralized data management: each microservice owns its own data/store to avoid a single database becoming a coupling/failure point.
Independent lifecycle: develop, test, deploy, scale and maintain services independently.
Scalability: scale individual services (for example, a payment service) rather than the whole application.
Fault isolation: failure of one microservice should not cascade to others.
Improved maintenance and reduced downtime: changes are localized to specific services.
Use of API gateways for routing and handling edge requests.

Real-world success stories and scale examples

Netflix: early adopter with ~500+ microservices; handles massive API traffic and performs very frequent deployments.
Amazon (AWS / Amazon shopping): a 2020 diagram cited ~23,000+ microservices; engineers deploy frequently (stat cited ~every 11.7 seconds).
Spotify: around ~810 services.
Uber: extensive use of many small services at large scale.

These examples illustrate operational scale, frequent deployments, and resilience benefits.

Essential technologies to learn (guide / checklist)

Containers
- Standard unit packaging an app plus all its dependencies (code, runtime, libraries, configuration).
- Ensures consistent behavior across environments (dev, SIT, UAT, prod).
- Typical pattern: one microservice → one container.
Docker
- Tool/platform to create container images.
- Dockerfile generates a read-only image; the Docker engine runs image instances as isolated containers.
- Typical workflow: package microservice into an image → run as a container → push to a registry → pull and run anywhere.
Kubernetes
- Open-source container orchestration system for managing many containers at scale.
- Automates deployment, scaling, and management of containerized applications in production.

Takeaway / practical guidance

To design and operate microservices, understand containers, how to build images with Docker, and how to orchestrate and scale them with Kubernetes.
Adopt decentralized data ownership, clear service boundaries, and automated CI/CD to enable frequent, safe deployments.