Friday, January 16, 2026
Microservices vs Monolith: The Split
Decide whether to break a monolithic Rails application into microservices for a rapidly scaling startup.
00The Situation
You are a Principal Engineer at a SaaS startup that has found product-market fit. The engineering team has grown from 10 to 50 engineers in the last year.
Current Architecture:
- Single monolithic Ruby on Rails application
- Postgres database (getting large, 2TB+)
- Redis for caching and background jobs (Sidekiq)
- Deployed on Kubernetes
The Problem:
- Deployments are slow (20+ minutes) and often break unrelated features
- Tests take 45 minutes to run in CI
- Different teams are stepping on each other's toes
- The database is becoming a bottleneck for some write-heavy features
The CTO is pushing to "move to microservices" to solve these velocity issues. She wants to split the app into 5-6 services immediately (Auth, Billing, Users, Notifications, Reporting, Core App).
Before responding, consider: What are the actual bottlenecks here? Are they technical or organizational?
Problem Analysis
5 min
Before proposing solutions, analyze whether microservices actually solve the stated problems.
Think about this first
Do you agree with the CTO's diagnosis? What questions would you ask?
Migration Strategy
15 min
If extraction is needed, design a safe migration path that doesn't halt feature development.
Think about this first
How would you migrate without stopping feature work or risking data integrity?
Operational Considerations
10 min
Microservices require significant operational investment. Address the hidden costs.
Think about this first
What operational challenges will the team face with microservices?