Skip to content


Switch branches/tags

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time

kafka-dropwizard-reporter Build Status

This package provides a DropwizardReporter class that connects the built-in metrics maintained by Kafka's client libraries with Dropwizard Metrics 3.0+. The Kafka metrics are added as Gauge instances to a Dropwizard MetricRegistry instance.

If you're already using Dropwizard Metrics in your application to serve metrics via HTTP, Graphite, StatsD, etc., this reporter provides an easy bridge to pass Kafka consumer, producer, and streams metrics to those same outputs.


dropwizard-metrics 3.0 and above.

kafka-clients 0.8.1 and above, including 0.9, 0.10, 0.11, and 1.0. Also functions with Kafka Streams and Kafka Connect.


First, declare a dependency on this package and on the explicit versions of the dependencies that you want:


      <!-- Required; you must provide metrics-core and kafka-clients;
           versions can be more recent than those shown below -->

      <!-- Optional; the user is expected to specify this dependency
           explicitly if DropwizardReporterGraphite is to be used -->

Then, include the DropwizardReporter class in the properties you pass to producers, consumers, and KafkaStreams applications:


That client will now automatically register all of its built-in metrics with a Dropwizard MetricRegistry when it's initialized. The registry is discovered by calling SharedMetricRegistries.getOrCreate("default"), so to direct DropwizardReporter to a particular registry, make sure to call SharedMetricRegistries.add("default", myRegistry) before instantiating Kafka clients if you want metrics to belong to myRegistry.

For a full example of integrating Kafka client metrics in a Dropwizard application, see example/.

Reporting to Graphite

If your application is not already handling reporting of metrics to an external source, you can use com.simple.metrics.kafka.DropwizardReporterGraphite which adds instantiation of a GraphiteReporter. Make sure you've declared a dependency on metrics-graphite, then see Configuration below.


The following configuration options are available:

Name Description Type Default Valid Values Importance
metric.dropwizard.graphite.hostDestination host for the GraphiteReporter (default: localhost); only relevant for DropwizardReporterGraphitestringlocalhostlow
metric.dropwizard.graphite.portDestination port for the GraphiteReporter (default: 2003); only relevant for DropwizardReporterGraphiteint2003low
metric.dropwizard.graphite.prefixMetric prefix for metrics published by the GraphiteReporter; only relevant for DropwizardReporterGraphitestring""low
metric.dropwizard.registryName of the dropwizard-metrics registry to use; passed to SharedMetricRegistries.getOrCreatestringdefaultlow

If you'd like to send Kafka metrics to a separate MetricRegistry instance, you can pass a name as metric.dropwizard.registry when configuring the client. For example, the following would end up calling SharedMetricRegistries.getOrCreate("kafka-metrics"):


A configuration using Graphite reporting could be:


Note that usage of these configuration options might trigger warning messages like org.apache.kafka.clients.consumer.ConsumerConfig: The configuration metric.dropwizard.registry = kafka-metrics was supplied but isn't a known config. due to a bug in Kafka's configuration machinery present until at least version This was addressed in KAFKA-3711.


To build the project, you'll need to install Apache Maven 3. If you're on Mac OS X, you can install Maven via Homebrew:

$ brew install maven

Once that's installed, run the following from main directory (where pom.xml lives):

mvn clean install


To deploy to Maven Central, you'll need to provide GPG signatures for all the artifacts. There's a sign profile for this purpose:

mvn clean verify -Psign

Check that the output looks good. You should a jar, sources, javadoc, and pom, each with a signed .asc companion:

$ ls target/kafka-*
target/kafka-dropwizard-reporter-1.1.1-javadoc.jar      target/kafka-dropwizard-reporter-1.1.1.jar
target/kafka-dropwizard-reporter-1.1.1-javadoc.jar.asc  target/kafka-dropwizard-reporter-1.1.1.jar.asc
target/kafka-dropwizard-reporter-1.1.1-sources.jar      target/kafka-dropwizard-reporter-1.1.1.pom
target/kafka-dropwizard-reporter-1.1.1-sources.jar.asc  target/kafka-dropwizard-reporter-1.1.1.pom.asc

If the target directory looks good, deploy:

mvn deploy -Psign

You'll need to have an account with Maven Central as part of the com.simple group. Contact the project maintainer for more info. Visit and close the new staging repository (named something like comsimple-XXXX). Wait a few minutes, then select the repo again and "Release" it. If there is no "Release" button, something is wrong with your artifacts, or they haven't been closed out yet.

Once released, it will take some time for Maven Central to sync and for the artifact to be available.


Contributions, feature requests, and bug reports are all welcome. Feel free to submit an issue or submit a pull request to this repo.


Check the releases page for summaries of what has changed in each release.

Also See

This project takes significant inspiration from the GraphiteReporter class defined in apakulov/kafka-graphite.


Register built-in Kafka client metrics to Dropwizard Metrics







No packages published