OpenMetadata/openmetadata-docs/content/v1.6.x-SNAPSHOT/connectors/ingestion/lineage/spark-lineage.md

title	slug
Spark Lineage Ingestion	/connectors/ingestion/lineage/spark-lineage

Spark Lineage Ingestion

A spark job may involve movement/transfer of data which may result into a data lineage, to capture such lineages you can make use of OpenMetadata Spark Agent which you can configure with your spark session and capture these spark lineages into your OpenMetadata instance.

In this guide we will explain how you can make use of the OpenMetadata Spark Agent to capture such lineage.

Requirement

To use the OpenMetadata Spark Agent, you will have to download the latest jar from here.

We support spark version 3.1 and above.

Configuration

While configuring the spark session, in this guide we will make use of PySpark to demonstrate the use of OpenMetadata Spark Agent

{% codePreview %}

{% codeInfoContainer %}

{% codeInfo srNumber=1 %}

Once you have downloaded the jar from here in your spark configuration you will have to add the path to your openmetadata-spark-agent.jar along with other required jars to run your spark job, in this example it is mysql-connector-java.jar

{% /codeInfo %}

{% codeInfo srNumber=2 %}

openmetadata-spark-agent.jar comes with a custom spark listener i.e. org.openmetadata.spark.agent.OpenMetadataSparkListener you will need to add this as extraListeners spark configuration.

{% /codeInfo %}

{% codeInfo srNumber=3 %}

spark.openmetadata.transport.hostPort: Specify the host & port of the instance where your OpenMetadata is hosted.

{% /codeInfo %}

{% codeInfo srNumber=4 %}

spark.openmetadata.transport.type is required configuration with value as openmetadata.

{% /codeInfo %}

{% codeInfo srNumber=5 %}

spark.openmetadata.transport.jwtToken: Specify your OpenMetadata Jwt token here. Checkout this documentation on how you can generate a jwt token in OpenMetadata.

{% /codeInfo %}

{% codeInfo srNumber=6 %}

spark.openmetadata.transport.pipelineServiceName: This spark job will be creating a new pipeline service of type Spark, use this configuration to customize the pipeline service name.

Note: If the pipeline service with the specified name already exists then we will be updating/using the same pipeline service.

{% /codeInfo %}

{% codeInfo srNumber=7 %}

spark.openmetadata.transport.pipelineName: This spark job will also create a new pipeline within the pipeline service defined above. Use this configuration to customize the name of pipeline.

Note: If the pipeline with the specified name already exists then we will be updating/using the same pipeline.

{% /codeInfo %}

{% codeInfo srNumber=8 %}

spark.openmetadata.transport.pipelineSourceUrl: You can use this configuration to provide additional context to your pipeline by specifying a url related to the pipeline.

{% /codeInfo %}

{% codeInfo srNumber=9 %}

spark.openmetadata.transport.pipelineDescription: Provide pipeline description using this spark configuration.

{% /codeInfo %}

{% codeInfo srNumber=10 %}

spark.openmetadata.transport.databaseServiceNames: Provide the comma separated list of database service names which contains the source tables used in this job. If you do not provide this configuration then we will be searching through all the services available in openmetadata.

{% /codeInfo %}

{% codeInfo srNumber=11 %}

spark.openmetadata.transport.timeout: Provide the timeout to communicate with OpenMetadata APIs.

{% /codeInfo %}

{% codeInfo srNumber=12 %}

In this job we are reading data from employee table and moving it to another table employee_new of within same mysql source.

{% /codeInfo %}

{% /codeInfoContainer %}

{% codeBlock fileName="pyspark.py" %}

from pyspark.sql import SparkSession

spark = (
    SparkSession.builder.master("local")
    .appName("localTestApp")

    .config(
        "spark.jars",
        "path/to/openmetadata-spark-agent.jar,path/to/mysql-connector-java-8.0.30.jar",
    )

    .config(
        "spark.extraListeners",
        "org.openmetadata.spark.agent.OpenMetadataSparkListener",
    )

    .config("spark.openmetadata.transport.hostPort", "http://localhost:8585")

    .config("spark.openmetadata.transport.type", "openmetadata")

    .config("spark.openmetadata.transport.jwtToken", "<openmetadata-jwt-token>")

    .config(
        "spark.openmetadata.transport.pipelineServiceName", "my_pipeline_service"
    )

    .config("spark.openmetadata.transport.pipelineName", "my_pipeline_name")

    .config(
        "spark.openmetadata.transport.pipelineSourceUrl",
        "http://your.org/path/to/pipeline",
    )

    .config(
        "spark.openmetadata.transport.pipelineDescription", "My ETL Pipeline"
    )

    .config(
        "spark.openmetadata.transport.databaseServiceNames",
        "random,local_mysql",
    )

    .config("spark.openmetadata.transport.timeout", "30")

    .getOrCreate()
)

# Read table using jdbc()

# Read from MySQL Table
employee_df = (
    spark.read.format("jdbc")
    .option("url", "jdbc:mysql://localhost:3306/openmetadata_db")
    .option("driver", "com.mysql.cj.jdbc.Driver")
    .option("dbtable", "employee")
    .option("user", "openmetadata_user")
    .option("password", "openmetadata_password")
    .load()
)

# Write data to the new employee_new table
(
    employee_df.write.format("jdbc")
    .option("url", "jdbc:mysql://localhost:3306/openmetadata_db")
    .option("driver", "com.mysql.cj.jdbc.Driver")
    .option("dbtable", "employee_new")
    .option("user", "openmetadata_user")
    .option("password", "openmetadata_password")
    .mode("overwrite")
    .save()
)

# Stop the Spark session
spark.stop()

{% /codeBlock %}

{% /codePreview %}

Once this pyspark job get finished you will see a new pipeline service with name my_pipeline_service generated in your openmetadata instance which would contain a pipeline with name my_pipeline as per the above example and you should also see lineage between the table employee and employee_new via my_pipeline.

{% image src="/images/v1.5/connectors/spark/spark-pipeline-service.png" alt="Spark Pipeline Service" caption="Spark Pipeline Service" /%}

{% image src="/images/v1.5/connectors/spark/spark-pipeline-details.png" alt="Spark Pipeline Details" caption="Spark Pipeline Details" /%}

{% image src="/images/v1.5/connectors/spark/spark-pipeline-lineage.png" alt="Spark Pipeline Lineage" caption="Spark Pipeline Lineage" /%}

Using Spark Agent with Databricks

Follow the below steps in order to use OpenMetadata Spark Agent with databricks.

1. Upload the jar to compute cluster

To use the OpenMetadata Spark Agent, you will have to download the latest jar from here and upload it to your databricks compute cluster.

To upload the jar you can visit the compute details page and then go to the libraries tab

{% image src="/images/v1.5/connectors/spark/spark-upload-jar.png" alt="Spark Upload Jar" caption="Spark Upload Jar" /%}

Click on the "Install Now" button and choose dbfs mode and upload the OpenMetadata Spark Agent jar.

{% image src="/images/v1.5/connectors/spark/spark-upload-jar-2.png" alt="Spark Upload Jar" caption="Spark Upload Jar" /%}

Once your jar is uploaded copy the path of the jar for the next steps.

{% image src="/images/v1.5/connectors/spark/spark-uploaded-jar.png" alt="Spark Upload Jar" caption="Spark Upload Jar" /%}

2. Create Initialization Script

Once your jar is uploaded you need to create a initialization script in your workspace.

#!/bin/bash

STAGE_DIR_JAR="<path to jar copied from step 1>"

echo "BEGIN: Upload Spark Listener JARs"
cp -f $STAGE_DIR_JAR /mnt/driver-daemon/jars || { echo "Error copying Spark Listener library file"; exit 1;}
echo "END: Upload Spark Listener JARs"

echo "BEGIN: Modify Spark config settings"
cat << 'EOF' > /databricks/driver/conf/openlineage-spark-driver-defaults.conf
[driver] {
  "spark.extraListeners" = "org.openmetadata.spark.agent.OpenMetadataSparkListener"
}
EOF
echo "END: Modify Spark config settings"

Note: The copied path would look like this dbfs:/FileStore/jars/.... you need to modify it like /dbfs/FileStore/jars/... this.

{% image src="/images/v1.5/connectors/spark/prepare-script.png" alt="Prepare Script" caption="Prepare Script" /%}

3. Configure Initialization Script

Once you have created a initialization script, you will need to attach this script to your compute instance, to do that you can go to advanced config > init scripts and add your script path.

{% image src="/images/v1.5/connectors/spark/prepare-script.png" alt="Prepare Script" caption="Prepare Script" /%}

{% image src="/images/v1.5/connectors/spark/spark-init-script.png" alt="Spark Init Script" caption="Spark Init Script" /%}

4. Configure Spark

After configuring the init script, you will need to update the spark config as well.

{% image src="/images/v1.5/connectors/spark/spark-config-set.png" alt="Spark Set Config" caption="Spark Set Config" /%}

these are the possible configurations that you can do, please refer the Configuration section above to get the detailed information about the same.

spark.extraListeners org.openmetadata.spark.agent.OpenMetadataSparkListener
spark.openmetadata.transport.type openmetadata
spark.openmetadata transport.pipelineSourceUrl http://<your-pipeline-host-port> 
spark.openmetadata transport.pipelineDescription "your pipeline description" 
spark.openmetadata.transport.hostPort https://<your-openmetadata-host-port> 
spark openmetadata transport.pipelineServiceName demo_pipeline 
spark.openmetadata transport.pipelineName demo_pipeline 
spark.openmetadata transport.databaseServiceNames db-service-name1,db-service-name2 
spark.openmetadata.transport.jwtToken <your-jwt-token> 
spark.openmetadata.transport.timeout 30

After all these steps are completed you can start/restart your compute instance and you are ready to extract the lineage from spark to OpenMetadata.