Which of the following will cause a Spark job to fail?

Question

Options

• ANever pulling any amount of data onto the driver node.
• BTrying to cache data larger than an executor's memory.
• CData needing to spill from memory to disk.
• DA failed worker node.
• EA failed driver node.

• A
  6% (3)
• C
  2% (1)
• D
  2% (1)
• E
  91% (49)

Explanation

The driver node is the central coordinator of a Spark application - it maintains the SparkContext, schedules all tasks, tracks their progress, and collects results. If the driver fails, the entire Spark application fails immediately with no automatic recovery. In contrast: worker/executor failures (option D) are tolerated because Spark can retry failed tasks on other executors using lineage; caching failures (option B) just mean the data will be recomputed on next access; data spilling to disk (option C) slows execution but does not cause failure; and never pulling data to the driver (option A) is actually a best practice, not a failure scenario.

No community discussion yet for this question.