1. What is the most reliable signal that a monolith should be split into microservices?
Microservices pay off when teams cannot ship independently. Tight coupling, long integration cycles, and a shared database that forces lockstep releases are strong signals. If a monolith still deploys quickly and safely, keep it. Optimize modularity first; split only where autonomy is worth the operational cost.
Symptom: one small change → rebuild whole app → coordinated release