Help Center/ Data Lake Insight/ User Guide/ Common DLI Management Operations/ Managing Program Packages of Jar Jobs/ Package Management Overview
Updated on 2024-11-08 GMT+08:00
View PDF
Share

Package Management Overview

Before running DLI jobs, UDF JAR files or Jar job packages need to be uploaded to the cloud platform for unified management and maintenance.

There are two ways to manage packages:

  • Upload packages to OBS: Upload Jar packages to an OBS bucket in advance and select the OBS path when configuring a job.
  • Upload packages to DLI: Upload Jar packages to an OBS bucket in advance, create a package on the Data Management > Package Management page of the DLI management console, and select the DLI package when configuring a job.

The package management function provided by the DLI console can help you manage various types of Jar job packages. This section explains how to upload and manage packages on the DLI management console.

  • For Flink 1.15 or later, only OBS packages can be selected when configuring jobs, and DLI packages are not supported.
  • For Spark 3.3.1 or later, you are advised to directly select OBS packages when configuring jobs.
  • When packaging Spark or Flink Jar jobs, do not upload the dependency packages that the platform already has to avoid conflicts with the built-in dependency packages of the platform. Refer to DLI Built-in Dependencies for built-in dependency packages.

Notes and Constraints

  • A package can be deleted, but a package group cannot be deleted.
  • The following types of packages can be uploaded:
    • JAR: JAR file
    • PyFile: User Python file
    • File: User file
    • ModelFile: User AI model file

Package Management Page

Table 1 Package management parameters

Parameter

Description

Group Name

Name of the group to which the package belongs. If the package is not grouped, -- is displayed.

Package Name

Name of a package.

Owner

Name of the user who uploads the package.

Type

Type of a package. The following package types are supported:

  • JAR: JAR file
  • PyFile: User Python file
  • File: User file
  • ModelFile: User AI model file

Status

Status of the package to be created.

  • Uploading: The file is being uploaded.
  • Finished: The resource package has been uploaded.
  • Failed: The resource package upload failed.

Created

Time when a package is created.

Updated

Time when the package is updated.

Operation

Manage Permissions: Manage user permissions for a package.

Delete: Delete the package.

More:

  • Modify Owner: Modify the owner of the package.
  • Tags: Add or edit package tags.