Answer-first summary for fast verification
Answer: The logic in the notebooks for Tasks A and B will be completed and committed successfully, but any operations performed by Task C might have been partially completed or remain partially persisted.
Databricks executes job tasks as independent units of work once their specific dependencies are met. Because Tasks A and B succeeded, their changes are committed to the Lakehouse. Databricks does not provide cross-task transactional atomicity; therefore, a failure in Task C does not trigger a rollback of Tasks A or B. Within Task C itself, while individual Delta Lake transactions are atomic (meaning an in-flight write would be rolled back), any writes that Task C completed before the failure or operations not wrapped in a single transaction may still persist. Key takeaway: Databricks Jobs provide failure isolation, not multi-task transactional integrity.
Author: LeetQuiz Editorial Team
Ultimate access to all questions.
A Databricks job is configured with three tasks: Task A runs independently, while Task B and Task C run in parallel once Task A completes. If Task A and Task B finish successfully but Task C encounters a failure, what is the resulting state of the data in the Lakehouse?
A
No changes will be saved to the Lakehouse unless all tasks are successful; the failure of Task C will trigger an automatic rollback of all changes made by Tasks A and B.
B
As tasks are managed based on dependencies, changes are held in a temporary state and won’t be committed to the Lakehouse until every task in the job finishes successfully.
C
The logic in the notebooks for Tasks A and B will be completed and committed successfully, but any operations performed by Task C might have been partially completed or remain partially persisted.
D
The logic in the notebooks for Tasks A and B will be completed successfully, but all changes made by Task C will be automatically undone by the Databricks Jobs service to maintain environment consistency.
No comments yet.