Python Micro-service Architecture
summary
A small Python shop migrates a monolith incrementally into single‑purpose services that own their data and communicate over REST the largest service is ~350 lines and deployments prove easier once APIs stabilize. Chosen tools include pyramid or flask for tiny HTTP services, requests for client calls, Python logging, tox, devpi for package hosting, pyenv, saltstack for config and a Python reimplementation of pact for consumer‑driven contract tests. Packaging and deployment remain work in progress but the approach already reduces coupling.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 1:09What are microservices?
- 2:13The problem with monoliths
- 3:24Considering solutions
- 5:07Why choose microservices?
- 7:19SOA vs microservices
- 8:39Implementation journey
- 11:14Future state goals
- 13:09Python tools for service
- 16:04Logging and debugging
- 17:29Testing and CI tools
- 19:14Managing Python versions
- 20:25Deployment management
- 21:16Contract testing with Pact
- 22:55Reflections on architecture
- 23:54Suggested reading
- 25:36Q&A: Scalability and dependencies
- 30:27Q&A: Deployment strategies
- 32:17Q&A: Advice for starting
related talks
30mHow To Build a Python Microservice Without Losing a Job
44mPython in the MicroServices world
33mLessons learned building Python microservices
25mA microservices architecture with Python and Docker containers
40mPython microservices on PaaS done right
37mEvolutionary Architecture & Microservices
18mMicroservices: the small print
29mPrototyping Python Microservices in Production