3-2-1 Backup Plan Generator
Start with a concrete topology.
Generate a planBuild independent recovery copies, size them, schedule them, and prove they can restore.
Start with a concrete topology.
Generate a planReserve version and snapshot capacity.
Plan retentionSet recovery expectations and test them.
Estimate recoveryRAID and snapshots are useful local tools, but independent copies remain the backup plan.
Planning path
Choose by failure question. The 3-2-1 generator maps copy roles; Local vs Cloud vs Hybrid chooses location mix; Retention sizes version history; Snapshot Storage estimates local rollback space; Offsite Capacity sizes remote destinations; Frequency selects recovery-point cadence; Recovery Time estimates restoration duration; Verification Schedule turns successful jobs into tested recovery evidence.
Define protected scope and acceptable data loss first. Select independent local and offsite roles, then size each destination with consistent retention. Check whether the initial seed and recurring jobs fit available transfer windows. Estimate full-restore time against the required recovery objective. Finish by assigning daily monitoring, sample restores, and recovery drills. A plan is incomplete until both capacity and tested procedure are owned.
Inventory critical data separately from replaceable data, average change rate, retention periods, destination independence, available bandwidth, effective restore speed, credential custody, encryption keys, owners, and the longest acceptable outage. Document which failure each copy covers. A second folder, snapshot, parity set, or replica under the same account may improve convenience without creating an independent recovery path.
Use Cloud Backup Upload & Restore Time Calculator for WAN feasibility and Backup Window Calculator for recurring changed data. Review the 3-2-1 guide, Snapshots vs Backups, and Local Backup vs Offsite Backup before finalizing policy. Put verification dates in an external calendar, retain restore evidence, and revise capacity, cadence, or topology whenever a test misses its recovery objective.