Surviving Microservices
summary
Argues that microservices work only if you invest up front: expect a quick POC in 6–12 weeks and a realistic go‑live around eight months, plus people, tooling and possibly consultants. Recommends local development with compose up, container images pushed to registries with explicit tags (never avoid latest in production), intentional logs and end‑to‑end tracing, message‑based eventual consistency with projectors or event sourcing for audit/history, repository-per-service and live failure drills before launch.
AI-written summary · about
chapters (19)
- 0:00Start
- 3:09What to Survive...
- 8:18Alignment of business capabilities/features
- 12:14Identifying Business Domains
- 13:32Identity Management Domain
- 22:30Full stack isolation
- 24:31HAD ONE JOB
- 25:20One microservice?
- 32:00Immutable images, and deployment
- 32:33One service, one source repo, one image repo
- 34:49Shared code - tsk tsk
- 37:29End to end tracing
- 38:14End to end visibility
- 41:02Group logs by microservice
- 43:20Eventual consistency
- 43:50Messaging-based solution
- 44:02Projecting data
- 44:45History / Audit
- 45:35Messaging and end-to-end visibility
related talks
52mSurviving Microservices
57mEmbracing Microservices
46mSurviving microservices
44mIs It Time Already? Microservices Are Dead; Long Live Microservices!
41mThe six obstacles of moving to microservices and how to deal with them
28mSurviving in a Microservices Environment
48mChallenges in Implementing MicroServices
45mBuilding and evolving microservices: lessons from the frontlines