Creating a plugin for elasticsearch 5.0 using Maven


Elasticsearch 5.0 switched to Gradle in October 2015.

You can obviously write a plugin using Gradle if you wish and you could benefit from all the goodies elasticsearch team wrote when it comes to integration tests and so on.

My colleague, Alexander Reelsen aka Spinscale on Twitter, wrote a super nice template if you wish to create an Ingest plugin for 5.0.

Hey! Wait! You wrote Ingest? What is that?

Ingest is a new feature coming in elasticsearch 5.0. It helps you to transform your data on the fly while injecting it into elasticsearch. Read more in elastic blog post.

If you know me and my work before I joined elastic, I have always been in love with data crawling and transformation as I wrote myself some plugins called rivers.

This blog post is part of a series which will teach you:

Let’s get started!

Create a skeleton

Create Maven skeleton

Create in your new project directory, let say ingest-bano, a pom.xml file:

<?xml version="1.0" encoding="UTF-8"?>
<project xmlns=""

    <name>Plugin: Ingest: BANO</name>
    <description>BANO Ingest Plugin for elasticsearch</description>


Add elasticsearch core dependency

Just add org.elasticsearch:elasticsearch:5.0.0 as a provided dependency:



Why provided? Actually our plugin will run in elasticsearch. So elasticsearch is already provided here.

Add plugin descriptor file

As explained in Plugin developer reference, you need to provide a which must be assembled with your plugin artifact in the elasticsearch dir.

Create this file in src/main/resources:


It does not need to be added to the project classes but only packaged within the ZIP file. Let’s tell Maven to do that!

We create a file src/main/assemblies/plugin.xml with:

<?xml version="1.0"?>

Note that we are filtering file which means that at package time all Maven placeholders will be replaced by their needed values.

We don’t want to add this file in our JAR so we tell that to Maven in the pom.xml:


We need to declare the maven assembly plugin now:

  <!-- ... -->

Configure Compiler plugin for Java 8

Because we want to use Java 8, we need to configure the compiler plugin:


Create Plugin class skeleton

Create an empty plugin class in src/main/java/org/elasticsearch/ingest/bano/.

package org.elasticsearch.ingest.bano;

import org.elasticsearch.plugins.Plugin;

public class IngestBanoPlugin extends Plugin {


Add test infrastructure

Elasticsearch provides a nice test framework which can be reused to run some tests.

Let’s add it to our project:


Let’s add the fantastic Randomized Testing framework:

    <!-- we skip surefire to work with randomized testing above -->

        <assertions enableSystemAssertions="false">

            <report-text />


Add a default log4j.xml in src/test/resources:

<?xml version="1.0" encoding="UTF-8" ?>
<!DOCTYPE log4j:configuration SYSTEM "log4j.dtd">

    <appender name="console" class="org.apache.log4j.ConsoleAppender">
        <layout class="org.apache.log4j.PatternLayout">
            <param name="ConversionPattern" value="[%d{ISO8601}][%-5p][%-25c] %m%n" />

        <level value="INFO" />
        <appender-ref ref="console" />


We can now extend ESIntegTestCase class to run integration tests. You can create a BanoPluginIntegrationTest class in src/test/java/org/elasticsearch/ingest/bano/:

package org.elasticsearch.ingest.bano;

import org.elasticsearch.plugins.Plugin;
import org.elasticsearch.plugins.PluginInfo;
import org.elasticsearch.test.ESIntegTestCase;

import java.util.Collection;
import java.util.Collections;

import static;

public class BanoPluginIntegrationTest extends ESIntegTestCase {

    protected Collection<Class<? extends Plugin>> nodePlugins() {
        return Collections.singleton(IngestBanoPlugin.class);

    public void testPluginIsLoaded() throws Exception {
        NodesInfoResponse response = client().admin().cluster().prepareNodesInfo().setPlugins(true).get();
        for (NodeInfo nodeInfo : response.getNodes()) {
            boolean pluginFound = false;
            for (PluginInfo pluginInfo : nodeInfo.getPlugins().getPluginInfos()) {
                if (pluginInfo.getName().equals(IngestBanoPlugin.class.getName())) {
                    pluginFound = true;
            assertThat(pluginFound, is(true));

When you run this class, a cluster of a random number of nodes is started using random Locale settings. The plugin we just created is added to nodes using nodePlugins() class.

So we just have to use the plugin as we want now. In this example, we are just testing that the plugin is actually loaded on every node.

Build the plugin

Just run:

mvn clean install

You will get in target/releases the distribution file.

Install the plugin

If you have an elasticsearch 5.0.0 version somewhere, you can install this plugin with:

bin/elasticsearch-plugin install file:///path/to/target/releases/

Then you can start elasticsearch with bin/elasticsearch and check in logs that the plugin is loaded.

You are now all set!


In a coming blog post, I’ll explain how to write an Ingest plugin based on the skeleton we just built and also how to create real integration tests.

But note that this skeleton can be used for whatever purpose:

Your imagination is now the limit! :)

Stay tuned!


David Pilato

20+ years of experience, mostly in Java. Living in Cergy, France.