Skip to main content

Java Registry

The Java registry implements MetricRegistry, wrapping the Prometheus Java library. This provides interoperability with anything that depends on the Java library.

ℹ️ The Java Registry does add a runtime constraint that goes beyond constraints that Prometheus itself imposes: You cannot have two metrics of the same name with different labels. This issue describes the problem.

See the example below on how to use the Java Registry:

import cats.effect.IO
import cats.effect.Resource

import io.prometheus.metrics.model.registry.PrometheusRegistry

import prometheus4cats.MetricFactory
import prometheus4cats.javaclient.JavaMetricRegistry

// Construct a Java registry using a default PrometheusRegistry
val default: Resource[IO, JavaMetricRegistry[IO]] =
  JavaMetricRegistry.Builder[IO]().build

// Construct a Java registry using a custom PrometheusRegistry
val custom: Resource[IO, JavaMetricRegistry[IO]] =
  JavaMetricRegistry.Builder[IO]().withRegistry(new PrometheusRegistry()).build

// Use the registry to get a factory
val factory: Resource[IO, MetricFactory[IO]] =
  custom.map(MetricFactory.builder.build(_))

Implementation Notes​

As per the MetricRegistry interface, this implementation returns Cats-Effect Resources to indicate a metric being registered and ultimately de-registered. Requesting a metric of the same name (and labels) multiple times will not result in an error, instead you will be returned the currently registered metric. The Java registry wrapper implements reference counting to ensure that a metric will only be de-registered when there are no more references to it or when the wrapper's surrounding Resource is finalized.