Service Architectures for Mere Mortals
summary
Argues that the move from Rails monolith to services hinges on concrete choices: keep one database per service and use Postgres if unsure; run background jobs with Sidekiq + Redis; prefer JSON/HTTP or lighter encodings like MessagePack or Protocol Buffers; use a Message Bus pub/sub for async notifications to avoid blocking HTTP. Warns about compensating actions, uptime stacking (99% × N), and shows that narrow services shrink CI/deploy to 29 seconds CI deploy in practice; demos Slack-as-bus but cuts the live demo.
AI-written summary · about
chapters (5)
- 0:00Introduction to SOA
- 8:54Monolith vs. Service
- 13:58Distributed best practices
- 21:26Communication patterns
- 30:45Trade-offs and conclusion
related talks
54mServices, Services, Everywhere!
27mScaling a Rails Monolith w/Event Driven Microservices
33mServices and Rails: The Sh*t They Don't Tell You
33mServices and Rails: The Shit They Don't Tell You
56mFrom Homogeneous Monolith to Heterogeneous Microservices Architecture
28mServices and Rails: The Shit They Don't Tell You
30mSome Funny Things Happened on The Way to A Service Ecosystem
37mArchitecture: The Next Generation