Python in the MicroServices world
summary
Built an all‑Python microservice (a simple lookup pulled out of a monolith) using Flask + Redis for metadata, then moved metadata into a separate store; deployment and image management prove painful so the team experiments with serverless and a message bus, later standardising on Kafka message bus and dead‑letter patterns. Authentication shifts to OpenID Connect; an action sometimes touches 25–30 services, teams of 4–5 own services, and the extracted lookup reduces DB load and cuts costs but exposes orchestration, tracing and CI/CD complexity.
AI-written summary · about
chapters (14)
- 0:00Intro
- 0:52About me
- 2:06MicroServices by Definition
- 3:24JEE EJB (Message Driven Beans)
- 13:05A typical MicroService
- 15:10Discipline
- 17:36Technical
- 20:26Business
- 23:55Organisational
- 30:25Distributed Transactions
- 33:08What happens to Sprints?
- 35:24Sealing the Contract
- 35:59What do we do?
- 37:46Branch and Merge Strategy
related talks
34mPython Micro-service Architecture
33mLessons learned building Python microservices
30mHow To Build a Python Microservice Without Losing a Job
25mA microservices architecture with Python and Docker containers
40mPython microservices on PaaS done right
29mPrototyping Python Microservices in Production
48mChallenges in Implementing MicroServices
49mChallenges in Implementing Microservices