A Databricks job has been configured with 3 tasks, each of which is a Databricks notebook. Task A does not depend on other tasks. Tasks B and C run in parallel, with each having a serial dependency…

Question

Options

• AAll logic expressed in the notebook associated with tasks A and B will have been successfully
• BAll logic expressed in the notebook associated with tasks A and B will have been successfully
• CAll logic expressed in the notebook associated with task A will have been successfully completed;
• DBecause all tasks are managed as a dependency graph, no changes will be committed to the
• EUnless all tasks complete successfully, no changes will be committed to the Lakehouse; because

• A
  87% (27)
• B
  3% (1)
• D
  3% (1)
• E
  6% (2)

Explanation

In a Databricks multi-task job, tasks execute and commit their results independently as they complete - there is no distributed transaction that rolls back other tasks if one fails. Task A ran first and completed successfully, committing all its notebook logic. Tasks B and C then ran in parallel; B completed successfully, committing all its notebook logic. When C failed, the work done by A and B had already been committed and persists in the Lakehouse. Only the work that was part of Task C's execution is incomplete or partially written depending on where in the notebook the failure occurred. The job run is marked as failed overall, but the successfully completed tasks (A and B) retain their committed state.

No community discussion yet for this question.