Datadog Metrics

This module provides a Metrics[F] with Events[F] trait that lets you send counters, histograms and events to the Datadog agent over UDP. Note that this means you are unlikely to receive errors if the Datadog agent isn’t reachable!

For more details about counters, histograms & the UDP format see the Datadog DogStatsD documentation. For more details about events see the event documentation.


In your build.sbt

libraryDependencies += "com.ovoenergy" %% "natchez-extras-dogstatsd" % "5.0.1"

Example usage

import cats.effect.{ExitCode, IO, IOApp}
import com.ovoenergy.natchez.extras.dogstatsd.Events.{AlertType, Priority}
import com.ovoenergy.natchez.extras.dogstatsd.{Dogstatsd, Events}
import com.ovoenergy.natchez.extras.metrics.Metrics
import com.ovoenergy.natchez.extras.metrics.Metrics.Metric

object MetricApp extends IOApp {

  val metricConfig: Dogstatsd.Config =

      // adds a prefix to all metrics, e.g. `my_app.metricname`
      metricPrefix = Some("my_app"),

      // the address your Datadog agent is listening on
      agentHost = SocketAddress(ip"", port"8125"),

      // these tags will be added to all metrics + events
      globalTags = Map("example_tag" -> "example_value")

  val exampleEvent: Events.Event =
      title = "Gosh, an event just ocurred!",
      body = "You should investigate this right away",
      alertType = AlertType.Warning,
      priority = Priority.Normal,
      tags = Map.empty

  val exampleCounter: Metric =
    Metric(name = "my_counter", tags = Map.empty)

  val exampleHistogram: Metric =
    Metric(name = "my_histogram", tags = Map.empty)

  def run(args: List[String]): IO[ExitCode] =
    Dogstatsd[IO](metricConfig).use { metrics: Metrics[IO] with Events[IO] =>
      for {
        _ <- metrics.counter(exampleCounter)(1)
        _ <- metrics.histogram(exampleHistogram)(1)
        _ <- metrics.event(exampleEvent)
      } yield ExitCode.Success