Production-Ready Microservices

Appendix A. Production-Readiness Checklist

This will be a checklist to run over all microservices—manually or in an automated way.

A Production-Ready Service Is Stable and Reliable

• It has a standardized development cycle.

• Its code is thoroughly tested through lint, unit, integration, and end-to-end testing.

• Its test, packaging, build, and release process is completely automated.

• It has a standardized deployment pipeline, containing staging, canary, and production phases.

• Its clients are known.

• Its dependencies are known, and there are backups, alternatives, fallbacks, and caching in place in case of failures.

• It has stable and reliable routing and discovery in place.

A Production-Ready Service Is Scalable and Performant

• Its qualitative and quantitative growth scales are known.

• It uses hardware resources efficiently.

• Its resource bottlenecks and requirements have been identified.

• Capacity planning is automated and performed on a scheduled basis.

• Its dependencies will scale with it.

• It will scale with its clients.

• Its traffic patterns are understood.

• Traffic can be re-routed in case of failures.

• It is written in a programming language that allows it to be scalable and performant.

• It handles and processes tasks in a performant manner.

• It handles and stores data in a scalable and performant way.

A Production-Ready Service Is Fault Tolerant and Prepared for Any Catastrophe

• It has no single point of failure.

• All failure scenarios and ...