The correct option is The system can be restored to normal operation after the outage.

A shutdown exercise is judged successful when it proves that recovery procedures actually return the environment to service. Demonstrating that The system can be restored to normal operation after the outage shows that backups, failover, configuration and dependency mapping worked and that recovery time objectives and service continuity can be met.

The outage causes minimal loss of data is not the primary indicator because low data loss relates to recovery point objectives and backup frequency. Minimal data loss can be important but it does not by itself show the system was fully recovered and operational.

The outage continues for the longest observed duration is wrong because a longer outage indicates failure not success. Duration is a measurement of impact and it should be minimized rather than maximized in a successful exercise.

The shutdown forces systems to crash is incorrect because forcing crashes can test robustness but causing crashes does not prove that controlled recovery and restoration procedures work. A successful exercise restores services cleanly rather than simply causing failures.

Full AWS Practitioner Certification Question