Microservices.md

Microservice Notes

Problems with a Large Codebase

• cognitive load
• unintended side effects
• releasing everything when you changed one thing
• scaling due to load on one component means you need to scale all components
• all must use the same framework/libraries/stack

How to mitigate these problems

• modular code
• clear boundaries within code for areas of responsibilities

Evolution of a Project (hardening):

Fancy graph showing size of project over product/market fit When to break out a microservice:

Fictitious Future System Diagram

cart before the horse?

Boundaries and Isolation

• Separating areas of concern
• Scaling components independently

Types of Boundaries

• Separate functions
• Python modules (django apps)
• Python packages/libraries/repos
• Microservices

Responsibilities of a Service

• Data in its area of concern
• Data stores (databases)
• Communication path to the service
• Scaling concerns
• transparency Logging/Alerting/Monitoring/Metrics (transparency of health of system)