Hitchhikers Guide to the Microservcies
summary
Argues that microservices solve specific problems but add complexity, so decide on them only after mapping dependencies and data sources; warns against polyglot stacks and recommends keeping languages minimal. Recommends investing in logging and metrics up front — ELK stack, Graphite/InfluxDB and Prometheus for metrics — and using Zipkin/OpenTracing for tracing. Describes deployment options (blue/green, canary, independent vs release-train) and notes a real extraction of a component and database that takes three months in one project. Emphasizes people and DevOps readiness.
AI-written summary · about
chapters (17)
- 0:00Introduction and welcome
- 1:10Speaker introduction
- 2:36Why microservices hype and risks
- 5:26Monoliths versus microservices
- 9:38Model it considerations
- 12:52Journey to microservices
- 17:01Splitting monoliths correctly
- 19:29Extracting services step by step
- 21:17Deployment strategies
- 27:13Troubleshooting in microservices
- 31:07Monitoring and tracing tools
- 33:59Cloud and tooling options
- 34:2012-factor app misconceptions
- 37:30Reactive manifesto basics
- 40:22Event sourcing and CQRS
- 42:29People and DevOps culture
- 44:57Summary and key takeaways
related talks
38mMicroservices, a bittersweet symphony
48mChallenges in Implementing MicroServices
52mMicroservices: A Utopian Mystery
50mChallenges in Implementing Microservices
43mPanel: Microservices, Where Did It All Go Wrong
37mKnee Deep in Microservices
39mConsidering Migrating a Monolith to Microservices?
45mResponsible Microservices