To diagnose spill in Spark, two primary indicators are found in the Stage’s detail screen and Executor’s log files. The Stage’s detail screen provides per-task metrics like 'Shuffle spill (memory)' and 'Shuffle spill (disk)', which directly indicate spill activity. The Executor’s log files may contain explicit messages about spilling data to disk. Other options are incorrect because spill metrics are task-specific (not in Query/Job/Driver details) and the Executor’s detail screen lacks per-partition spill data.