Topic guide

Architecture & Scaling

Architecture decisions with explicit trade-offs: monolith or microservices, when to scale, technical debt, rewrites, and build vs. buy.

3 topics · 12 articles

Architecture patterns

Proven patterns and how to record the decision.

Scaling decisions

When and how to scale systems without losing reliability.

Technology trade-offs

Choosing the right tool and dealing with legacy.

Frequently asked questions

When should we move to microservices?

After proving the monolith is the bottleneck, whether for deployment, scale, or coordination between teams. That usually shows up around 30 engineers, not earlier.

Monolith or microservices for a startup?

A well-modularized monolith. Microservices add operational cost (deployment, observability, consistency) that a startup does not need to pay yet.

How many microservices is too many?

When there are more services than people able to be on call for them. For most mid-size companies, that lands between 5 and 15.

When do we need a data warehouse?

When reports and analytics start slowing down the transactional (OLTP) database. Before that, a read replica usually does the job.