Which statement best describes the purpose of a Kubernetes Job?

Which statement best describes the purpose of a Kubernetes Job?

• A. A Job runs indefinitely and never terminates.
• B. A Job ensures a specified number of pods terminate successfully.
• C. A Job creates pods that always restart after termination.
• D. A Job is used to expose services externally.

Answer: (B)

A Kubernetes Job is designed for running a finite, batch task and ensuring it finishes. It creates pods to perform the work and tracks how many of those pods complete successfully. You specify how many completions you want, and the Job keeps creating pods (including retrying failed ones) until that many pods have finished. Once the required number of successful completions is reached, the Job is considered done. This makes it perfect for one-off data processing, ETL, or any task that must run to completion rather than stay running. The other scenarios describe behaviors that don’t fit a Job’s purpose: running indefinitely is typical of a long-running service, not a batch job; pods that always restart after termination describe an always-on or restart-policy scenario that isn’t about completing a set number of tasks; and exposing services externally is the domain of Service or Ingress resources, not a Job.

A Kubernetes Job is designed for running a finite, batch task and ensuring it finishes. It creates pods to perform the work and tracks how many of those pods complete successfully. You specify how many completions you want, and the Job keeps creating pods (including retrying failed ones) until that many pods have finished. Once the required number of successful completions is reached, the Job is considered done. This makes it perfect for one-off data processing, ETL, or any task that must run to completion rather than stay running.

The other scenarios describe behaviors that don’t fit a Job’s purpose: running indefinitely is typical of a long-running service, not a batch job; pods that always restart after termination describe an always-on or restart-policy scenario that isn’t about completing a set number of tasks; and exposing services externally is the domain of Service or Ingress resources, not a Job.