Step 1: Effect of a crash during recovery.
If a database system crashes while recovery is in progress and
no checkpointing is used, the system has no saved recovery state.
As a result, when the system restarts, the recovery procedure must begin again from the start by scanning the log.
Step 2: Handling of undo and redo operations.
In the absence of checkpoints, the system cannot know which undo or
redo actions were already completed before the second crash.
Therefore, the same undo list and redo list are reconstructed from the log and applied again during recovery.
Step 3: Evaluate the given options.
Option (A): Correct, because recovery restarts from the beginning with the same undo and redo lists.
Option (B): Incorrect, since recovery is still possible even without checkpointing.
Option (C): Incorrect, because operations may indeed be repeated when there is no checkpoint.
Option (D): Incorrect, as standard recovery mechanisms ensure that database consistency is eventually restored.
Final Conclusion:
The correct statement is (A).