Jobs¶

A Job is a workload object used to manage one-time tasks, ensuring that a specified number of Pods run successfully until the task is completed (rather than running continuously). It is suitable for batch processing jobs, data processing tasks, or scenarios that require reliable execution and termination. If a Pod fails (e.g., due to program errors or node failures), the Job automatically restarts or recreates the Pod until the task reaches the preset number of successful completions.

For more information, refer to Jobs.

List¶

Go to Infrastructure > Containers, click Kubernetes > Jobs in the upper left corner of the page to view information about all Jobs retained in the workspace, including the job name, number of parallel tasks, number of running Pods, running duration, etc.

Details¶

Click on a data entry in the list to slide out its details page.

Basic Attributes¶

View the number of parallel tasks, running duration, labels, and extended attributes of the Job.

• Click Parallel Tasks/Running Duration to perform the following operations:

• Labels

Automatically uploaded with Job information by default. Existing labels can be displayed in the Job list through quick filters to show data with the same labels.