Real Metrics from Our Microservices

32 custom metrics tracking everything from CRUD operations to page render times.

32 Custom Metrics

Complete Coverage

Our implementation includes 32 custom metrics across both services:

Backend Service: 21 metric beans
Frontend Service: 11 metric beans

These metrics provide complete visibility into business operations, not just infrastructure.

Backend Service: 21 Metrics

CRUD Operations (Counters)

person.operations.total{operation="create", result="success"}
person.operations.total{operation="create", result="error"}
person.operations.total{operation="read", result="success"}
person.operations.total{operation="update", result="success"}
person.operations.total{operation="update", result="error"}
person.operations.total{operation="delete", result="success"}
person.operations.total{operation="delete", result="error"}

7 metrics tracking every CRUD operation with success/error split

Service Execution Time (Timers)

person.service.execution.time{method="create"}
person.service.execution.time{method="findById"}
person.service.execution.time{method="findAll"}
person.service.execution.time{method="update"}
person.service.execution.time{method="delete"}

5 metrics measuring performance of service methods

Repository & REST API (Mixed)

person.repository.total (Gauge - current count)
person.api.requests.total{endpoint="/people", method="GET"}
person.api.requests.total{endpoint="/people", method="POST"}
person.api.requests.total{endpoint="/people/{id}", method="GET"}
person.api.requests.total{endpoint="/people/{id}", method="PUT"}
person.api.requests.total{endpoint="/people/{id}", method="DELETE"}
...and 3 more validation/error metrics

9 metrics for database and API tracking

Frontend Service: 11 Metrics

Frontend Operations (Counters)

person.frontend.requests.total{operation="list", result="success"}
person.frontend.requests.total{operation="view", result="success"}
person.frontend.requests.total{operation="create", result="success"}
person.frontend.requests.total{operation="create", result="error"}
person.frontend.requests.total{operation="update", result="success"}
person.frontend.requests.total{operation="delete", result="success"}

6 metrics tracking user interactions

Backend Communication (Counters)

person.frontend.backend.errors.total{type="connection"}
person.frontend.backend.errors.total{type="timeout"}
person.frontend.backend.errors.total{type="server_error"}

3 metrics tracking backend communication failures

Page Render Performance (Timers)

person.frontend.page.render.time{page="list"}
person.frontend.page.render.time{page="form"}

2 metrics measuring UI performance

The MetricsConfig Pattern

Centralized Bean Configuration

All metrics are defined as @Bean methods in a centralized MetricsConfig class. This pattern provides several benefits:

Centralized: All metrics in one place
Reusable: Inject via @Autowired + @Qualifier
Testable: Easy to mock in unit tests
Type-Safe: Compile-time validation
Documented: Bean names serve as documentation

Example: Creating a Counter Bean

@Configuration
class MetricsConfig {
    @Bean
    Counter personCreateCounter(MeterRegistry registry) {
        return registry.counter("person.operations.total",
                "operation", "create", "result", "success")
    }
}

Example: Using the Counter

@Service
class PersonService {
    @Autowired
    @Qualifier("personCreateCounter")
    Counter personCreateCounter

    def create(Person person) {
        // ... business logic ...
        personCreateCounter.increment()
        return person
    }
}

Why Track Business Metrics?

📊 Business Value

Infrastructure metrics (CPU, memory) tell you about resources. Business metrics tell you about actual user behavior and business outcomes.

🎯 Better Decisions

"How many people are we creating per minute?" is more actionable than "CPU is at 60%". Business metrics guide product decisions.

🔍 Faster Debugging

"Create operation error rate spiked" is more specific than "Error logs increased". Pinpoint exactly which operation failed.

📈 Trend Analysis

Track growth over time. Are we handling more creates today than yesterday? Last week? Last month? Business metrics show trends.

Key Takeaways

32 custom metrics provide complete coverage (21 backend + 11 frontend)
Centralized MetricsConfig pattern for easy management
Track business operations, not just infrastructure
Metrics injected via Spring dependency injection with @Qualifier
Every CRUD operation, service method, and API endpoint tracked