A Tale of Microservices
summary
A five-person team splits a product into microservices, builds a data pipeline and a data‑visualization service that reads Google Sheets, using Python and Node for appropriate components. They containerize everything with Docker (optimizing Docker layer order to cut CI builds from ~10 minutes to ~2–3 minutes), deploy with Kubernetes and Helm, and monitor with Prometheus, Grafana and the ELK stack; tracing revealed an 8‑second stall that was fixed by moving a heavy job to a fork.
AI-written summary · about
chapters (18)
- 0:00Microservice architecture intro
- 5:48What is microservice architecture
- 8:17Unix philosophy and microservices
- 9:19Data visualization demo
- 11:42Splitting monolith into microservices
- 12:47Service naming and responsibility
- 14:12Tech stack selection
- 16:17Development processes and CI/CD
- 18:48Optimizing Docker builds
- 21:16Microservice testing and integration
- 23:47Monitoring distributed systems
- 27:23Infrastructure and deployment
- 28:30Kubernetes for service management
- 31:01Scaling challenges and dependencies
- 32:51Logging and error analysis
- 34:17Tracing requests and latency
- 36:06When microservices are needed
- 36:50Summary
related talks
25mA microservices architecture with Python and Docker containers
36mCultivating a Microservice Culture Via Tooling
01h 01mMicroservices Gone Wrong
01h 07mMicroservices from the Trenches. A Guide through the Architecture, Design, Code
35mSeven years with microservices: a critical review
37mKnee Deep in Microservices
35mMicroservices: A Tale of 3 Heroes
51mMicroservices and Microfrontends: Scaling, Learning, Growing