Setting the User Context

Describes how to set the user context when using the Spark OSS images.

User context is not configured automatically for the Spark OSS images.

If you do not set the user context, the user identity that is set in the image is used to access the data sources like persistent volumes. This can cause your Spark application to fail due to a lack of proper permissions.

To set the logged-in user’s context when using the Spark OSS images, add the following configuration in the sparkConf section of your Spark application YAML file.
spark.hpe.webhook.security.context.autoconfigure: "true"

Once you add this configuration to the YAML file, HPE Ezmeral Unified Analytics Software automatically defines and sets the user context for the logged-in user, and your Spark application runs successfully.

NOTE
Do not add this configuration to your Spark application YAML file if you are using the default Spark images that support Data Fabric services.

To learn more about pod security context, see Configure a Security Context for a Pod or Container.