# Micrometer — Application Metrics Facade for the JVM

> Micrometer is a vendor-neutral application observability facade for the JVM that provides a simple API for collecting metrics and exporting them to monitoring systems like Prometheus, Datadog, and Grafana.

## Install

Save as a script file and run:

# Micrometer — Application Metrics Facade for the JVM

## Quick Use
```java
// Add dependency: io.micrometer:micrometer-registry-prometheus:1.13.0
MeterRegistry registry = new PrometheusMeterRegistry(PrometheusConfig.DEFAULT);
Counter requestCounter = registry.counter("http.requests", "method", "GET");
requestCounter.increment();
Timer.Sample sample = Timer.start(registry);
// ... do work ...
sample.stop(registry.timer("http.request.duration", "endpoint", "/api"));
```

## Introduction
Micrometer is the metrics collection facade for JVM-based applications, often called the SLF4J of metrics. It provides a vendor-neutral API for recording application metrics — counters, gauges, timers, distribution summaries, and histograms — and ships them to the monitoring backend of your choice. Micrometer is the default metrics library in Spring Boot Actuator.

## What Micrometer Does
- Provides a unified API for recording counters, gauges, timers, and distribution summaries in JVM applications
- Exports metrics to 15+ monitoring systems including Prometheus, Datadog, New Relic, InfluxDB, and CloudWatch
- Tracks JVM internals automatically: memory usage, garbage collection, thread pools, and class loading
- Records HTTP server and client metrics with automatic dimensional tagging for Spring Boot applications
- Supports histogram and percentile computation for latency distribution analysis

## Architecture Overview
Micrometer defines a MeterRegistry abstraction that each monitoring backend implements. Application code records metrics through Meter objects (Counter, Timer, Gauge, DistributionSummary) bound to a registry. Tags (key-value pairs) add dimensions to every metric for filtering and grouping. Composite registries allow publishing the same metrics to multiple backends simultaneously. In Spring Boot, auto-configuration creates the appropriate registry bean and binds default JVM and HTTP metrics automatically.

## Self-Hosting & Configuration
- Add the core `micrometer-core` dependency plus the registry module for your monitoring backend
- In Spring Boot, add `micrometer-registry-prometheus` and metrics are automatically exposed at `/actuator/prometheus`
- Configure common tags in `application.properties` to add environment, service name, or region labels to all metrics
- Customize metric names and tags using `MeterFilter` beans for renaming, dropping, or transforming meters
- Set histogram bucket boundaries and percentile targets for timer and distribution summary metrics

## Key Features
- Vendor-neutral API — switch monitoring backends by changing a dependency, not application code
- Dimensional metrics with tags enable flexible filtering, grouping, and aggregation in dashboards
- Automatic instrumentation for Spring MVC, WebFlux, JDBC, Kafka, and cache libraries
- Rate aggregation and histogram support built into timers and distribution summaries
- Active maintenance by the Spring team with releases aligned to Spring Boot versions

## Comparison with Similar Tools
- **Prometheus Java Client** — Direct Prometheus instrumentation without the abstraction layer; lower overhead but locks you into one backend
- **Dropwizard Metrics** — Predecessor metrics library with codahale-style naming; lacks dimensional tagging and multi-backend support
- **OpenTelemetry Java SDK** — Broader observability scope covering traces, metrics, and logs, but heavier and more complex for metrics-only use cases
- **StatsD** — UDP-based metrics aggregation protocol; simple but requires a separate aggregation daemon and lacks rich type support

## FAQ
**Q: Is Micrometer only for Spring Boot?**
A: No. Micrometer is a standalone library that works with any JVM application. Spring Boot provides auto-configuration, but you can create and manage MeterRegistries manually in any Java, Kotlin, or Groovy project.

**Q: Can I send metrics to multiple monitoring systems?**
A: Yes. Use a CompositeMeterRegistry to publish the same metrics to Prometheus, Datadog, and any other supported backend simultaneously.

**Q: How does Micrometer relate to OpenTelemetry?**
A: Micrometer focuses on metrics with a clean, type-safe API. Micrometer Tracing (formerly Spring Cloud Sleuth) bridges to OpenTelemetry for distributed tracing. You can use both together.

**Q: Does Micrometer add significant overhead?**
A: Micrometer is designed for production use with minimal overhead. Meter lookups are cached, and recording operations are lock-free where possible.

## Sources
- https://github.com/micrometer-metrics/micrometer
- https://micrometer.io/

---
Source: https://tokrepo.com/en/workflows/asset-5e10f8fe
Author: Script Depot