version: "1.0.0"

Skill: Microservices & Distributed Systems

Usage: Use when designing or interacting with scalable, independent service boundaries.

🕸️ Core Philosophy

• Independent Deployability: A single microservice can be changed and deployed independently of the rest of the system.
• Loose Coupling: Services should know as little about each other as possible.
• High Cohesion: Related behavior should be grouped together.

📡 Communication Patterns

• Synchronous (REST/gRPC): Use when an immediate response is required. Be wary of tight coupling and cascading failures.
• Asynchronous (Pub/Sub, Event Streaming): Prefer async communication using message brokers (RabbitMQ, Kafka, Azure Service Bus) for high availability and decoupling.
• Event-Driven Architecture: Services react to events emitted by other services rather than direct calls.

🗄️ Data Management

• Database per Service: Each microservice must have its own private database to ensure loose coupling. Never share a DB.
• Saga Pattern: Manage distributed transactions across multiple services using a sequence of local transactions, coordinated by a central orchestrator or choreography.
• CQRS (Command Query Responsibility Segregation): Separate the models used for updating data (Commands) from the models used for reading data (Queries).
• Event Sourcing: Store the state of a domain entity as a sequence of state-changing events.

🛡️ Resilience & Fault Tolerance

• Circuit Breaker Pattern: Prevent failures in a downstream service from cascading upstream by quickly failing calls to an unhealthy service.
• Retry with Exponential Backoff: Handle transient networking issues gracefully.
• Bulkheads: Isolate resources (like connection pools) for different services to prevent one failing service from exhausting all resources.
• Timeouts: Always set timeouts on network calls.

🚫 Anti-Patterns

• Distributed Monolith: Services are deployed separately but are tightly coupled (e.g., sharing a database, synchronous chatty communication).
• Building microservices without robust automated deployment and monitoring in place.