How to Back Up a Digital Local-History Archive So One Drive Failure Cannot Destroy It

An external drive is not an archive strategy when it is the only copy.

One failed disk, accidental deletion, ransomware event or lost account should not be able to erase years of preservation work.

Define the authoritative archive

Identify what actually belongs in the protected master set.

That may include:

• preservation masters;
• metadata;
• transcripts;
• OCR;
• indexes or databases;
• checksum manifests;
• README files;
• rights and provenance records.

Do not carefully back up thumbnails while forgetting the masters.

Use the 3-2-1 model

A practical pattern is:

• 3 copies
• 2 different storage media or systems
• 1 offsite

For example:

1. working archive on the main computer or NAS;
2. local external backup;
3. independent offsite or cloud copy.

Three folders on one drive are still one failure domain.

Make copies fail differently

Backups should not all disappear from the same event.

Think about:

• fire;
• theft;
• flood;
• ransomware;
• mistaken deletion;
• account lockout;
• lightning.

Two external drives beside the same computer improve redundancy but do not provide much disaster separation.

Keep one copy offline or otherwise protected

A permanently mounted backup can be reached by some of the same malware or mistakes that damage the working archive.

Useful options include:

• disconnected external drive;
• versioned cloud storage;
• immutable/object-lock storage;
• rotated offline media.

The correct choice depends on archive size and budget.

Treat sync and backup differently

Cloud sync is useful.

It can also synchronize deletion and corruption.

Check:

• version history;
• deleted-file retention;
• account recovery;
• MFA;
• export capability.

A sync client can be one layer without being the entire backup strategy.

Generate checksums

Checksums help detect unexpected changes.

SHA-256 manifests are a practical modern option.

For preservation masters:

1. generate a manifest;
2. copy it with the archive;
3. preserve a copy independently;
4. periodically recompute;
5. compare;
6. investigate mismatches.

A checksum detects change.

It does not create another copy.

Test restoration

Periodically restore a sample from the independent backup.

Open files.

Verify checksums.

Restore metadata or the archive database.

A backup that has never been restored is an assumption.

Keep versions

Do not overwrite one backup generation forever.

Keep older known-good points, especially before:

• major reorganization;
• mass renaming;
• metadata migration;
• format conversion.

That gives you somewhere to return if an error propagates.

Plan for encryption and succession

Private archives may contain living-person information.

Encrypt offsite copies when appropriate.

Then make sure trusted successors can recover the encryption key if the archive is meant to outlive its current custodian.

An encrypted archive nobody can decrypt is not preserved.

Refresh aging storage

Hard drives and flash media are not permanent shelves.

Periodically:

• power and test drives;
• review SMART/health indicators where useful;
• migrate to newer media;
• verify file counts and checksums.

Do not expect one USB drive to remain trustworthy for decades without inspection.

Document the system

Keep a simple README describing:

• authoritative master location;
• copy locations;
• backup method;
• schedule;
• checksum method;
• restore procedure;
• responsible person.

Automation is useful, but failed automation must be visible.

The goal is not “a backup job ran.”

The goal is “the archive can be recovered.”

For collection organization, see How to Organize a Large Family Photograph Collection Digitally.