Just Enough Distribution (JED) on Monoliths and Microservices

The post argues that the debate between monolithic architecture and microservices is often framed too extremely. It says microservices can bring benefits but also add complexity, while calling monoliths the only right choice is also an oversimplification. The better choice depends on the project’s context, performance needs, constraints, team expertise, cost, reliability, fault tolerance, and time to market.

The author prefers a middle-ground approach: start with a monolith and split out specific functionalities into microservices when needed. The term JED, or Just Enough Distribution, is presented as the best way to describe that approach.

Reading notes

#

• Rejects the idea that microservices automatically solve bad architecture.
• Says monolithic architecture is not the only valid approach.
• Lists context, performance requirements, project constraints, team expertise, cost constraints, reliability and fault tolerance, and time to market as factors in the decision.
• Prefers starting with a monolith and breaking out specific functionalities into microservices.
• Uses JED, Just Enough Distribution, for that middle-ground view.