Skip to main content

Package Tasks (Flink/Spark Jar & PySpark)

Package task development primarily focuses on running custom Jar tasks. Users can upload custom Jar packages, configure task parameters, and set runtime resource parameters.

  1. Click Data Intelligence Development > Package Tasks in the left navigation to enter the package task management page, where you can see all package tasks under the current workspace along with historical versions of each task.

  1. Click the Add Task button, and in the popup interface, set the task name, processing engine, select an application package, set the main class, parameters, and configuration items.

Processing engines include Flink, Spark, and Ray. After selecting Spark, you can choose Spark App Jar type or PySpark mode; after selecting Ray, you can choose a Ray Runtime Env package as the runtime environment.

  1. After adding a task, an Apply for Launch button appears in the right action bar. Click it to submit the task for launch approval.

  1. Task Update: Click the edit button in the action bar to publish a new version of the current task. After updating, a new Apply for Launch button appears for another launch application.

  1. After applying for launch, the task appears in the Production Tasks > Publish Approval interface where administrators can review and process it.