Which parameter must be published in your custom transformer to enable parallel processing?

Which parameter must be published in your custom transformer to enable parallel processing?

• A. Group By Parameter
• B. Output Port
• C. Reader
• D. WorkspaceRunner

Answer: (A)

Parallel processing of a custom transformer relies on partitioning the incoming features into independent groups that can be processed at the same time. To enable this, you publish a parameter that defines how features are grouped—the Group By Parameter. This parameter tells the engine which attribute(s) define a group, so each group can run concurrently on separate threads. Without a published group-by parameter, there’s no defined way to split work, so processing stays serial. The other options don’t control this parallel partitioning: the Output Port directs where results go, the Reader brings data in, and the WorkspaceRunner runs an entire workspace rather than enabling per-transform parallelism.

Parallel processing of a custom transformer relies on partitioning the incoming features into independent groups that can be processed at the same time. To enable this, you publish a parameter that defines how features are grouped—the Group By Parameter. This parameter tells the engine which attribute(s) define a group, so each group can run concurrently on separate threads. Without a published group-by parameter, there’s no defined way to split work, so processing stays serial. The other options don’t control this parallel partitioning: the Output Port directs where results go, the Reader brings data in, and the WorkspaceRunner runs an entire workspace rather than enabling per-transform parallelism.