Breaking the monolith with Node and Docker at Netflix
summary
Netflix migrates a Java UI data API monolith into many Node services running in Docker, claiming per-service runtime isolation, independent scaling, and the same operational dashboards and tracing as before. The team builds a Node platform, dev tooling that hot-reloads containers so edits appear in seconds instead of minutes, and CI/CD plus local debugging. Metrics: 1000 data endpoint scripts consolidated into isolated services handled by 200 UI developers, validated in production using shadow traffic and telemetry.
AI-written summary · about
chapters (23)
- 0:00Intro
- 0:42About Netflix
- 1:27Java REST API Server
- 3:39Serverless Infrastructure
- 4:39Inefficient Developer Workflow
- 7:07Java Data API Server
- 7:45Slay the Monolith
- 11:27Use Restify
- 12:04Route Management
- 13:18Connect Style Middleware
- 15:17Require Expire
- 16:30Simplify Environment Set-up
- 17:02Streamline Workflow with Tooling
- 19:18Efficient Workflow
- 21:04Java API Server
- 21:41Operational Visibility
- 22:14Infrastructure Integration : Atlas Request/Response Count
- 23:05Server Restart Count
- 23:17CPU and Heap Usage
- 23:31Query Server Log Real Time
- 24:10Distributed Request Tracing
- 25:08Shadow Test in Production
- 26:46Goal Reached
related talks
22mRapidly Developing Microservices with Node js and Docker
50mFast Delivery
30mApplication Modernization Using Cloud Strangler Pattern and DevOps Best Practices
31mKeynote Panel: Node.js at Massive Scale
29mNode.js Micro-Services
28mScaling Node.js Applications with Kubernetes and Docker
25mNode.js Meets Docker
22mObservable Node.js Applications