Developing and Testing ETL Scripts Locally Using the AWS Glue ETL Library

The AWS Glue Scala library is available in a public Amazon S3 bucket, and can be consumed by the Apache Maven build system. This enables you to develop and test your Python and Scala extract, transform, and load (ETL) scripts locally, without the need for a network connection.

Local development is available for all AWS Glue versions, including AWS Glue version 0.9 and AWS Glue version 1.0 and later. For information about the versions of Python and Apache Spark that are available with AWS Glue, see the Glue version job property.

The library is released with the Amazon Software license (https://aws.amazon.com/asl).

Note

The instructions in this section have not been tested on Microsoft Windows operating systems.

For local development and testing on Windows platforms, see the blog Building an AWS Glue ETL pipeline locally without an AWS account

Local Development Restrictions

Keep the following restrictions in mind when using the AWS Glue Scala library to develop locally.

• Avoid creating an assembly jar ("fat jar" or "uber jar") with the AWS Glue library because it causes the following features to be disabled:

  • Job bookmarks

  • AWS Glue Parquet writer (format="parquet")

  • FillMissingValues transform (Scala or Python)

  These feature are available only within the AWS Glue job system.

• The FindMatches transform is not supported with local development.

Developing Locally using Docker image

Docker image gives you a two-step process to set up a container with AWS Glue binaries and a Jupyter/Zeppelin notebook server. For more information, see Developing AWS Glue ETL jobs locally using a container.

Developing Locally with Python

Complete some prerequisite steps and then use AWS Glue utilities to test and submit your Python ETL script.

Prerequisites for Local Python Development

Complete these steps to prepare for local Python development:

1. Clone the AWS Glue Python repository from GitHub (https://github.com/awslabs/aws-glue-libs).

2. Do one of the following:

   • For AWS Glue version 0.9, check out branch glue-0.9.

   • For AWS Glue versions 1.0, check out branch glue-1.0. All versions above AWS Glue 0.9 support Python 3.

   • For AWS Glue versions 2.0, check out branch glue-2.0.

   • For AWS Glue version 3.0, check out the master branch.

3. Install Apache Maven from the following location: https://aws-glue-etl-artifacts.s3.amazonaws.com/glue-common/apache-maven-3.6.0-bin.tar.gz.

4. Install the Apache Spark distribution from one of the following locations:

   • For AWS Glue version 0.9: https://aws-glue-etl-artifacts.s3.amazonaws.com/glue-0.9/spark-2.2.1-bin-hadoop2.7.tgz

   • For AWS Glue version 1.0: https://aws-glue-etl-artifacts.s3.amazonaws.com/glue-1.0/spark-2.4.3-bin-hadoop2.8.tgz

   • For AWS Glue version 2.0: https://aws-glue-etl-artifacts.s3.amazonaws.com/glue-2.0/spark-2.4.3-bin-hadoop2.8.tgz

   • For AWS Glue version 3.0: https://aws-glue-etl-artifacts.s3.amazonaws.com/glue-3.0/spark-3.1.1-amzn-0-bin-3.2.1-amzn-3.tgz

5. Export the SPARK_HOME environment variable, setting it to the root location extracted from the Spark archive. For example:

   • For AWS Glue version 0.9: export SPARK_HOME=/home/$USER/spark-2.2.1-bin-hadoop2.7

   • For AWS Glue version 1.0 and 2.0: export SPARK_HOME=/home/$USER/spark-2.4.3-bin-spark-2.4.3-bin-hadoop2.8

   • For AWS Glue version 3.0: export SPARK_HOME=/home/$USER/spark-3.1.1-amzn-0-bin-3.2.1-amzn-3

Running Your Python ETL Script

With the AWS Glue jar files available for local development, you can run the AWS Glue Python package locally.

Use the following utilities and frameworks to test and run your Python script. The commands listed in the following table are run from the root directory of the AWS Glue Python package.

Utility	Command	Description
AWS Glue Shell	./bin/gluepyspark	Enter and run Python scripts in a shell that integrates with AWS Glue ETL libraries.
AWS Glue Submit	./bin/gluesparksubmit	Submit a complete Python script for execution.
Pytest	./bin/gluepytest	Write and run unit tests of your Python code. The pytest module must be installed and available in the PATH. For more information, see the pytest documentation.

Developing Locally with Scala

Complete some prerequisite steps and then issue a Maven command to run your Scala ETL script locally.

Prerequisites for Local Scala Development

Complete these steps to prepare for local Scala development.

Step 1: Install Software

In this step, you install software and set the required environment variable.

1. Install Apache Maven from the following location: https://aws-glue-etl-artifacts.s3.amazonaws.com/glue-common/apache-maven-3.6.0-bin.tar.gz.

2. Install the Apache Spark distribution from one of the following locations:

   • For AWS Glue version 0.9: https://aws-glue-etl-artifacts.s3.amazonaws.com/glue-0.9/spark-2.2.1-bin-hadoop2.7.tgz

   • For AWS Glue version 1.0: https://aws-glue-etl-artifacts.s3.amazonaws.com/glue-1.0/spark-2.4.3-bin-hadoop2.8.tgz

   • For AWS Glue version 2.0: https://aws-glue-etl-artifacts.s3.amazonaws.com/glue-2.0/spark-2.4.3-bin-hadoop2.8.tgz

   • For AWS Glue version 3.0: https://aws-glue-etl-artifacts.s3.amazonaws.com/glue-3.0/spark-3.1.1-amzn-0-bin-3.2.1-amzn-3.tgz

3. Export the SPARK_HOME environment variable, setting it to the root location extracted from the Spark archive. For example:

   • For AWS Glue version 0.9: export SPARK_HOME=/home/$USER/spark-2.2.1-bin-hadoop2.7

   • For AWS Glue version 1.0 and 2.0: export SPARK_HOME=/home/$USER/spark-2.4.3-bin-spark-2.4.3-bin-hadoop2.8

   • For AWS Glue version 3.0: export SPARK_HOME=/home/$USER/spark-3.1.1-amzn-0-bin-3.2.1-amzn-3

Step 2: Configure Your Maven Project

Use the following pom.xml file as a template for your AWS Glue Scala applications. It contains the required dependencies, repositories, and plugins elements. Replace the Glue version string with one of the following:

• 3.0.0 for AWS Glue version 3.0

• 1.0.0 for AWS Glue version 1.0 or 2.0

• 0.9.0 for AWS Glue version 0.9

  <project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/maven-v4_0_0.xsd">
    <modelVersion>4.0.0</modelVersion>
    <groupId>com.amazonaws</groupId>
    <artifactId>AWSGlueApp</artifactId>
    <version>1.0-SNAPSHOT</version>
    <name>${project.artifactId}</name>
    <description>AWS ETL application</description>

        <properties>
        <scala.version>*<2.11.1> for AWS Glue 2.0 or below, or <2.12.7> for AWS Glue 3.0*</scala.version>
        </properties>

    <dependencies>
        <dependency>
            <groupId>org.scala-lang</groupId>
            <artifactId>scala-library</artifactId>
            <version>${scala.version}</version>
        </dependency>
        <dependency>
            <groupId>com.amazonaws</groupId>
            <artifactId>AWSGlueETL</artifactId>
			
            <version>Glue version</version>
			
        </dependency>
    </dependencies>

    <repositories>
        <repository>
            <id>aws-glue-etl-artifacts</id>
            <url>https://aws-glue-etl-artifacts.s3.amazonaws.com/release/</url>
        </repository>
    </repositories>
    <build>
        <sourceDirectory>src/main/scala</sourceDirectory>
        <plugins>
            <plugin>
                <!-- see http://davidb.github.com/scala-maven-plugin -->
                <groupId>net.alchim31.maven</groupId>
                <artifactId>scala-maven-plugin</artifactId>
                <version>3.4.0</version>
                <executions>
                    <execution>
                        <goals>
                            <goal>compile</goal>
                            <goal>testCompile</goal>
                        </goals>
                    </execution>
                </executions>
            </plugin>
            <plugin>
                <groupId>org.codehaus.mojo</groupId>
                <artifactId>exec-maven-plugin</artifactId>
                <version>1.6.0</version>
                <executions>
                    <execution>
                        <goals>
                        <goal>java</goal>
                        </goals>
                    </execution>
                </executions>
                <configuration>
                <systemProperties>
                    <systemProperty>
                        <key>spark.master</key>
                        <value>local[*]</value>
                    </systemProperty>
                    <systemProperty>
                        <key>spark.app.name</key>
                        <value>localrun</value>
                    </systemProperty>
                    <systemProperty>
                        <key>org.xerial.snappy.lib.name</key>
                        <value>libsnappyjava.jnilib</value>
                    </systemProperty>
                </systemProperties>
                </configuration>
            </plugin>
            <plugin>
                <groupId>org.apache.maven.plugins</groupId>
                <artifactId>maven-enforcer-plugin</artifactId>
                <version>3.0.0-M2</version>
                <executions>
                    <execution>
                        <id>enforce-maven</id>
                        <goals>
                            <goal>enforce</goal>
                        </goals>
                        <configuration>
                            <rules>
                                <requireMavenVersion>
                                    <version>3.5.3</version>
                                </requireMavenVersion>
                            </rules>
                        </configuration>
                    </execution>
                </executions>
            </plugin>
        </plugins>
    </build>
</project>

Running Your Scala ETL Script

Run the following command from the Maven project root directory to run your Scala ETL script.

mvn exec:java -Dexec.mainClass="mainClass" -Dexec.args="--JOB-NAME jobName"

Replace mainClass with the fully qualified class name of the script's main class. Replace jobName with the desired job name.

Configuring a Test Environment

For examples of configuring a local test environment, see the following blog articles:

• Building an AWS Glue ETL pipeline locally without an AWS account

• Developing AWS Glue ETL jobs locally using a container

If you want to use development endpoints or notebooks for testing your ETL scripts, see Developing Scripts Using Development Endpoints.

Note

Development endpoints are not supported for use with AWS Glue version 2.0 jobs. For more information, see Running Spark ETL Jobs with Reduced Startup Times.