Skip to main content Skip to complementary content

Enabling lineage for Big Data Jobs

With certain Hadoop distributions, you can trace how your big data is handled through the input, the processing and the output steps of a Talend Big Data Batch Job.

Setting up data lineage with Cloudera Navigator

The support for Cloudera Navigator has been added to Talend Spark Jobs.

If you are using Cloudera V5.5+ to run your Jobs, you can make use of Cloudera Navigator to trace the lineage of given data flow to discover how this data flow was generated by a Spark Job, including the components used in this Job and the schema changes between the components.

If you are using CDP Private Cloud Base or CDP Public Cloud to run your Jobs, it is recommended to use Apache Atlas. If you are using CDP dynamic distribution, Apache Atlas is used rather than Cloudera Navigator. For more information, see Setting up data lineage with Atlas.

For example, assume that you have designed the following Job and you want to generate lineage information about it:

Spark Jobs running.
  1. Click Run to open its view and then click the Hadoop configuration tab. For a Spark Job, the tab to be used is Spark configuration.
  2. From the Distribution list, select Cloudera and from the Version list, select Cloudera 5.5. Then the Use Cloudera Navigator check box is displayed.

    With this option activated, you need to set the following parameters:

    • Username and Password: this is the credentials you use to connect to your Cloudera Navigator.

    • Cloudera Navigator URL: enter the location of the Cloudera Navigator to be connected to.

    • Cloudera Navigator Metadata URL: enter the location of the Navigator Metadata.

    • Activate the autocommit option: select this check box to make Cloudera Navigator generate the lineage of the current Job at the end of the execution of this Job.

      Since this option actually forces Cloudera Navigator to generate lineages of all its available entities such as HDFS files and directories, Hive queries or Pig scripts, it is not recommended for the production environment because it will slow the Job.

    • Kill the job if Cloudera Navigator fails: select this check box to stop the execution of the Job when the connection to your Cloudera Navigator fails. Otherwise, leave it clear to allow your Job to continue to run.

    • Disable SSL validation: select this check box to make your Job to connect to Cloudera Navigator without the SSL validation process.

      This feature is meant to facilitate the test of your Job but is not recommended to be used in a production cluster.

Till now, the connection to Cloudera Navigator has been set up. The time when you run this Job, the lineage will be automatically generated in Cloudera Navigator.

Note that you still need to configure the other parameters in the Spark configuration tab in order to successfully run the Job.

When the execution of the Job is done, perform a search in Cloudera Navigator for the data written by this Job and see the lineage of this data in Cloudera Navigator.

If you compare this lineage graph with the Job in Talend Studio, you can see that every component is presented in this graph and you can expand the icon of each component to read the schema it uses.

Lineage graph in Cloudera Navigator.

Cloudera Navigator uses a Cloudera SDK library to provide functionalities and must be compatible with the version of this SDK library. The version of your Cloudera Navigator is determined by the Cloudera Manager installed with your Cloudera distribution and the compatible SDK is automatically used based on the version of your Navigator.

However, not all the Cloudera Navigator versions have their compatible SDK versions. For more details about the Cloudera SDK versions and their compatible Navigator versions, see the Cloudera documentation about Cloudera Navigator SDK Version Compatibility.

For information about Cloudera Navigator versions supported by Talend Studio, see Supported Cloudera Navigator versions for Talend Jobs.

Setting up data lineage with Atlas

The support for Apache Atlas has been added to Talend Spark Jobs.

If you are using Hortonworks Data Platform V2.4 onwards to run your Jobs and Apache Atlas has been installed in your Hortonworks cluster, you can make use of Atlas to trace the lineage of given data flow to discover how this data was generated by a Spark Job, including the components used in this Job and the schema changes between the components. If you are using CDP Private Cloud Base or CDP Public Cloud to run your Jobs and Apache Atlas has been installed in your cluster, you can also make use of Atlas.

Depending on the Hortonworks Data Platform version you are using, Talend Studio supports the following Atlas version for:
  • Hortonworks Data Platform V2.4, the Talend Studio supports Atlas 0.5 only.
  • Hortonworks Data Platform V2.5, the Talend Studio supports Atlas 0.7 only.
  • Hortonworks Data Platform V3.14, the Talend Studio support Atlas 1.1 only.

For example, assume that you have designed the following Spark Batch Job and you want to generate lineage information about it in Atlas:

Spark Batch Job opened.

In this Job, tRowGenerator is used to generate the input data, tMap and tSortRow to process the data and the other components to output data into different formats.

  1. Click Run to open its view and then click the Spark configuration tab.
  2. From the Distribution list and the Version list, select your Hortonworks distribution. The Use Atlas check box is displayed.

    With this option activated, you need to set the following parameters:

    • Atlas URL: enter the location of the Atlas to be connected to. It is often http://name_of_your_atlas_node:port

    • In the Username and Password fields, enter the authentication information for access to Atlas.

    • Set Atlas configuration folder: if your Atlas cluster contains custom properties such as SSL or read timeout, select this check box, and in the displayed field, enter a directory in your local machine, then place the atlas-application.properties file of your Atlas in this directory. This way, your Job is enabled to use these custom properties.

      You need to ask the administrator of your cluster for this configuration file. For further information about this file, see the Client Configs section in Atlas configuration.

    • Die on error: select this check box to stop the Job execution when Atlas-related issues occur, such as connection issues to Atlas. Otherwise, leave it clear to allow your Job to continue to run.

Till now, the connection to Atlas has been set up. The time when you run this Job, the lineage will be automatically generated in Atlas.

Note that you still need to configure the other parameters in the Spark configuration tab in order to successfully run the Job. For further information, see Creating Spark Batch Jobs.

When the execution of the Job is done, perform a search in Atlas for the lineage information written by this Job and read the lineage there.

Reading the Atlas lineage

In Atlas, the lineage written by a Job consists of two types of entities:

    • The Job itself.

    • The components in the Job that are using data schemas, such as tRowGenerator or tSortRow. The connection or configuration components such as tHDFSConfiguration are not taken into account since these components do not use schemas.

The example Job generates 6 entities: one for the Job and five for the components, and automatically adds three different tags to these entities:

  • Talend for all the entities generated by the Job

  • TalendComponent for all the component entities.

  • TalendJob for all the Job entities.

You can directly click one of these tags in Atlas to display the corresponding entities.

For example, you can click TalendJob to show a list of Job entities logged in Atlas and select a Job to see the lineage information of this Job. The following image shows how the data flow is handled in the Job used as example:
Lineage graph in Atlas.

Did this page help you?

If you find any issues with this page or its content – a typo, a missing step, or a technical error – please let us know!

Leave your feedback here