To guard against data loss when the hard disk of a server fails, you should:
-
create a copy of the System State data.
-
create two sets of backups.
-
install the Recovery Console as a startup option.
-
keep the installation CD at an easily accessible location.
Creating two sets of backups provides redundancy - if one backup set is damaged, corrupted, or lost, you have another set available for recovery. System State data alone (A) is insufficient for full disaster recovery. Recovery Console (C) helps repair systems but doesn't protect against data loss. Installation media (D) is needed but doesn't constitute a backup strategy.
Creating two sets of backups is correct. The listed distractors don't actually protect data on a failed physical disk: System State backup only covers OS/registry/AD-type data, not all files; the Recovery Console is a boot-time repair tool, not a data-protection measure; and keeping the install CD accessible only helps you reinstall the OS, not recover lost data. The only option that directly guards against permanent data loss from a hard-disk failure is having backups — and specifically maintaining two independent backup sets (e.g., on separate media/rotations) protects against the case where one backup set itself is corrupt, incomplete, or unreadable at restore time, which is the standard redundancy recommendation in server administration best practice.