Skip to content

Spring Boot Microservices — Basics You'll Need (Cheat Sheet)

A practical, beginner-to-intermediate checklist for building Spring Boot microservices with production concerns: configuration, resilience, observability, deployment.

1) What makes a service a microservice?

  • Single responsibility / bounded context (domain-driven boundary)
  • Independently deployable unit
  • Owns its data (avoid shared DB)
  • Communicates via well-defined APIs (HTTP/gRPC) and/or events (Kafka/Rabbit)

Dependencies (typical): - spring-boot-starter-web (REST) - spring-boot-starter-validation - spring-boot-starter-actuator (health/metrics) - micrometer-registry-prometheus (if Prometheus) - spring-boot-starter-security (if needed) - spring-boot-starter-data-jpa + DB driver (if relational) - Resilience: spring-cloud-starter-circuitbreaker-resilience4j - Tracing: micrometer-tracing-bridge-otel (Boot 3+)

3) API design essentials

DTOs & validation

record CreateOrderRequest(
  @jakarta.validation.constraints.NotBlank String customerId,
  @jakarta.validation.constraints.Positive int quantity
) {}

Controller best practices

  • Keep controllers thin (delegate to service layer)
  • Use @ControllerAdvice for consistent error responses
  • Use versioning (/v1/...) or header-based versioning

4) Data patterns you’ll actually need

Database-per-service

  • Each service owns its schema.
  • Cross-service queries happen via APIs or events.

Transactions across services (don’t do 2PC)

  • Prefer Saga pattern (choreography via events or orchestration).
  • Use Outbox pattern to publish events reliably from DB.

5) Resilience patterns (must know)

Circuit breaker + timeout + bulkhead

import io.github.resilience4j.circuitbreaker.annotation.CircuitBreaker;
import org.springframework.stereotype.Service;

@Service
class PricingClient {
  @CircuitBreaker(name = "pricing", fallbackMethod = "fallback")
  String getPrice(String sku) {
    // call downstream
    throw new RuntimeException("downstream failed");
  }

  String fallback(String sku, Throwable t) {
    return "PRICE_UNAVAILABLE";
  }
}
Rules of thumb: - Timeouts everywhere (client & server). - Retries only for safe/idempotent ops. - Bulkheads to avoid thread/connection pool exhaustion.

6) Observability (logs + metrics + traces)

Actuator & health

management.endpoints.web.exposure.include=health,info,metrics,prometheus
management.endpoint.health.show-details=when_authorized

Structured logging

  • Include traceId / spanId in logs (log correlation).

Tracing (OpenTelemetry)

  • Use Micrometer Tracing bridge (Boot 3+) or OTel Java agent.

7) Configuration & secrets

  • Use env vars / config server, never hardcode secrets.
  • Externalize config per environment: dev/test/prod.

8) Security basics

  • Prefer OAuth2/OIDC with JWT for service-to-service.
  • Validate tokens at the edge (gateway) AND at services for defense-in-depth.

9) Testing pyramid for microservices

  • Unit tests (fast)
  • Slice tests (@WebMvcTest, @DataJpaTest)
  • Contract tests (consumer-driven contracts)
  • Integration tests with Testcontainers

10) Deployment basics

  • Containerize (Docker)
  • Health probes: readiness/liveness via Actuator
  • Resource limits (CPU/memory)
  • Rolling deployments + backward compatible APIs

Common interview questions

  • Why database-per-service?
  • How do you do distributed transactions?
  • What are timeouts/retries/circuit breakers and how do you tune them?
  • How do you debug latency across services?

References (official docs to bookmark)

  • Spring Boot Actuator docs
  • Spring Cloud Circuit Breaker docs
  • OpenTelemetry Java instrumentation docs