Distributed Services

Tracing

Especially in distributed services, tracing is significant to maintain a system. Requests are sent across multiple microservices. Latency, identifying correlating requests and aggregating the date across multiple services is essential for analysis.

• Spring Cloud Sleuth: Implements distributed tracing in Spring Boot appliactions.
• Zipkin or Jaeger: Zipkin or Jaeger are distributed systems for visualizing trace data within and between services.

TraceId, spanId gets propagated from one service to another via request headers.

Health check

Spring Boot Actuator enables creating custom health checks for monitoring application health. It's rather easy to implement a custom actuator to collect data to show in your monitoring endpoint. Spring Boot has a number of built-in actuator endpoints out of the box. e.g. metrics

@Component
public class MyHealthIndicator extends AbstractHealthIndicator {
    @Override
    protected void doHealthCheck(Health.Builder builder) throws Exception {
        boolean isHealthy = checkSystem();
        if (isHealthy) {
            builder.up().withDetail("Database", "Available");
        } else {
            builder.down().withDetail("Database", "Not Available");
        }
    }
}

Service Discovery

With the Eureka framework, it is rather easy to set up a service discovery server and clients.

@SpringBootApplication
@EnableEurekaServer
public class EurekaServerApplication {
    public static void main(String[] args) {
        SpringApplication.run(EurekaServerApplication.class, args);
    }
}

@SpringBootApplication
@EnableEurekaClient
public class AccountsServiceApplication {
    public static void main(String[] args) {
        SpringApplication.run(AccountsServiceApplication.class, args);
    }
}

Further information

Link: Service discovery with spring boot

Retryable api calls

The @Retryable annotation allows retrying a method call if it fails. If you experience e.g. network timeouts, it can improve the reliability of your service.

@Retryable(value = {IOException.class}, maxAttempts = 5, backoff = @Backoff(delay = 2000))
public String fetchData() {
    // API call that might fail
}

• maxAttemps: Number of attempts before giving up
• backoff: time between retries