RAID and Backups Are Not the Same Thing

Two different problems

When people first build a storage box, a common assumption is that using several drives in a RAID arrangement means their data is safe. That is only partly true, and the gap can be costly. RAID and backups are meant to answer different questions.

• RAID answers: “If one drive dies, can the system keep working?” It is about availability.
• A backup answers: “If something goes wrong with my data, can I get an earlier copy back?” It is about recovery.

Mixing them up is one of the most common ways home lab users lose files.

What RAID does

RAID combines several physical drives into one logical volume. The details differ by level, and exact names and behaviour can vary between implementations, but the general ideas are:

• Striping spreads data across drives for speed or capacity. On its own it adds no protection, and losing any one drive typically loses the whole volume. That makes it riskier than a single drive.
• Mirroring keeps identical copies on two or more drives. If one fails, another still holds everything.
• Parity stores extra calculated information so that the contents of a failed drive can be reconstructed from the others. Some layouts tolerate one failed drive and others tolerate two.

The benefit is that a single drive failure need not interrupt you. You replace the drive, the array rebuilds, and in the meantime the files remain reachable. For a server that other things depend on, that is valuable.

What RAID does not protect against

RAID copies whatever is written, including mistakes. It does not hold earlier versions. Common situations where RAID gives no help:

• Accidental deletion. Delete a folder and the deletion is applied to every drive at once.
• Overwriting or editing a file wrongly. The previous version is gone from all drives.
• File system corruption. A software bug, a bad update or an unclean shutdown can damage the structures that describe your files. The damage is stored redundantly, too.
• Malware and ransomware. Software that encrypts or wipes files does so across the whole volume. If the storage is mounted on an infected computer, the encrypted version replaces the originals.
• Controller or software faults. A failing controller, or a configuration mistake, can affect the whole array rather than one drive.
• Fire, flood, power surges and theft. All drives in one box share the same fate if the box is destroyed or taken.
• Multiple failures. Drives bought together and used together can wear at similar rates. A second drive may fail during a rebuild, especially with large drives where rebuilds are lengthy. How likely this is depends on the drives and layout, so do not assume it will not happen.
• Mistakes during recovery. Replacing the wrong drive in a degraded array is a classic way to turn a small problem into a large one.

What a backup is

A backup is a separate copy of data, stored somewhere that a problem with the original does not automatically reach. Several properties make a copy worth calling a backup:

• Separate. It lives on different hardware, ideally in a different place.
• Versioned or point-in-time. Ideally you can reach older states, so that a mistake noticed days later can still be undone.
• Not constantly writable from the source. Copies that a compromised computer can overwrite or encrypt are weaker. Offline or write-protected copies resist this better.
• Tested. A backup you have never restored is a hope, not a plan.

The usual guideline for how many copies, on what media and where is covered in the discussion of the 3-2-1 idea for hobby data, which builds on this distinction.

Syncing services and mirrored folders deserve a note. A tool that keeps two locations identical copies deletions and corruption as faithfully as it copies good files, so by itself a sync is closer to RAID than to a backup, unless it keeps previous versions.

Why people still use RAID

Redundancy is useful for the right job. If a machine hosts services that others use, or if rebuilding from backup would take a long time, RAID reduces downtime. It also spares you the disruption of restoring everything after one ordinary drive failure. It simply should not be the only protection. A sensible setup often combines the two: RAID for continuity, backups for recovery.

Some other points when planning a home NAS:

• Drives of similar age are likely to wear similarly, so consider staggering purchases or monitoring drive health reports.
• Keep a record of which physical drive sits in which bay, so replacement is not a guess.
• Know how your chosen layout rebuilds and what happens if a second drive fails mid-rebuild.
• Decide how much usable space you are willing to give up for redundancy, since parity and mirroring both reduce it.

A quick self-check

Ask yourself these questions about any important files:

1. If I deleted this by accident last Tuesday, could I get it back?
2. If the storage box were stolen, burned or flooded, would a copy remain?
3. If ransomware encrypted my computer and its mounted shares, is there a copy it cannot reach?
4. Have I actually restored from the backup recently, to confirm it works?

If any answer is no, RAID alone is not covering that risk. Add a separate backup, test it, and treat the array as a convenience that keeps the machine running rather than as the safe place where the data is kept. Outcomes with storage are never guaranteed, which is exactly why having more than one kind of protection is the goal.