During a citywide log-parse MapReduce job, one mapper node crashes mid-task. What resilience behavior should operators expect in a classic MapReduce framework?
Select an answer to reveal the explanation.
Short Explanation
If one oven goes cold, move those trays to another oven—don't rebuild the kitchen. MapReduce retries failed tasks on healthy workers. It does not reformat the cluster or magically rewrite the job as Spark streaming.
Full Explanation
Fault tolerance in MapReduce centers on re-scheduling failed tasks onto other workers while preserving successful progress elsewhere when possible. A single mapper crash should not require reformatting the cluster or aborting every completed reduce output by policy. Frameworks also do not auto-convert a MapReduce job definition into a Spark streaming pipeline.