The 3-2-1 Backup Rule: Three Copies, Two Media, One Offsite

• Also see: The Model Context Protocol explained
• Also see: How to build a PKM system that works
• Also see: How to lock down your online accounts
• Also see: How to wire your home network

The 3-2-1 backup rule is the entire strategy in one sentence: keep three copies of your data, on two different media, with one copy offsite. It is the most widely recommended data protection scheme in existence, and almost nobody actually does it. The reason is not that it is complicated. It is that it requires thinking about loss before it happens, and humans are very good at not thinking about loss until it is too late.

Why three copies

The first copy is your live data: the files on your laptop, the photos on your phone. This is the copy you work with every day, and the copy most likely to be lost, because it lives on a device that travels, gets dropped, gets stolen, and eventually fails.

The second copy is your local backup: a complete duplicate on a separate device in the same location. A NAS, an external hard drive, a second computer. When your laptop drive fails at 11 p.m., the local backup means you are back up and running by lunch the next day.

The third copy is your offsite backup: a complete duplicate stored somewhere else. A cloud service, a drive at a friend’s house, a safe deposit box. If your house burns down, the local backup burns with it. The offsite copy survives.

Why two media

Different failure modes affect different technologies. Hard drives fail mechanically. SSDs fail electronically. Cloud services have their own problems, including the provider going bankrupt or losing your data through error.

If all three copies are on the same type of media, a single systemic failure can take out all three. A firmware bug that bricks a particular SSD model. A ransomware strain that targets a particular NAS operating system. Media diversity ensures that no single technological failure can destroy every copy of your data.

In practice: your live data on an internal SSD, your local backup on a NAS with hard drives, your offsite backup in the cloud. Three copies, three different storage technologies, three different failure profiles.

Why one offsite

A fire that destroys your laptop also destroys the NAS sitting next to it. A thief who steals your laptop may also steal the external drive in the same bag. Ransomware that encrypts your laptop can spread to every device on the same network.

Geographic separation breaks these correlations. A copy that is not in your house cannot be destroyed by a fire in your house. A copy that is not on your network cannot be encrypted by ransomware on your network. A copy that is not in your bag cannot be stolen with your bag.

The failure modes

Drive failure is the most common. Any single drive failure destroys one copy. With three copies, you lose one and have two left. Theft destroys the live copy and potentially the local backup if they are in the same location. The offsite copy is the only thing that survives. Fire destroys everything in the house; only the offsite copy survives.

Ransomware is the newest and most insidious. It encrypts every file it can reach, including files on connected drives and network shares. A NAS on the same network as an infected laptop can be encrypted along with the laptop. The offsite copy, if it is a cloud backup with versioning, survives because it is not continuously connected to your network and because you can roll back to a pre-infection version.

flowchart TD
    A["A data loss event occurs"] --> B{"What was lost?"}
    B -->|"Single drive failure"| C["Restore from local NAS"]
    B -->|"Theft or fire"| D["Restore from cloud backup"]
    B -->|"Ransomware"| E["Restore from cloud backup with versioning"]
    C --> F{"Restore verified?"}
    D --> F
    E --> F
    F -->|"Yes"| G["Back to work"]
    F -->|"No"| H["Try the other copy"]
    H --> F

That flowchart is the backup decision tree. Different failures require different recovery paths, and the 3-2-1 rule ensures you always have a path available.

Implementation: NAS plus cloud

The simplest implementation for most people is a local NAS plus a cloud backup service.

For the local NAS, a two-bay Synology or QNAP with two hard drives in a mirrored configuration gives you redundancy against a single drive failure. Set it up to back up your laptop and desktop on a schedule, nightly or continuous. The NAS becomes your second copy: always available, always current, always local.

For the cloud backup, Backblaze Personal Backup, iDrive, or a similar service provides unlimited or large-capacity backup for a flat annual fee. Install the client on your laptop and desktop, point it at your important folders, and let it run. The cloud becomes your third copy: offsite, versioned, and independent of your local hardware.

The total cost is roughly the price of the NAS plus drives, plus about $100 per year for the cloud service. For most people, that is less than the cost of a single data recovery attempt, which typically runs $500 to $3,000 and has no guarantee of success.

Backup is not sync

This distinction is critical and widely misunderstood. Sync services like Dropbox, Google Drive, and iCloud are not backups. If you delete a file on one device, the deletion syncs to all devices. If ransomware encrypts a file, the encrypted version syncs to all devices.

A backup is a point-in-time copy that does not change unless you tell it to. A good backup service keeps multiple versions of your files, so you can restore not just the current state but the state from last week or last month. This versioning is what makes ransomware recovery possible: you roll back to a version from before the infection.

The rule of thumb: if deleting a file on your laptop deletes it everywhere, you have sync, not backup. If deleting a file on your laptop leaves the backup copy intact, you have a backup.

Testing restores

The most common backup mistake is never testing a restore. A backup that has never been restored is not a backup; it is a hypothesis. It might be corrupted, incomplete, or silently failing for months. You will not know until you need it, and by then it is too late.

Test your restores quarterly. Pick a random folder from your local NAS and restore it to a test location. Verify the files open and are not corrupted. Then do the same from your cloud backup. This takes about 30 minutes and is the only way to know your backup actually works.

Common mistakes

Never testing is the first mistake. Single points of confusion is the second: having multiple backup systems that overlap in confusing ways, so you are never sure which copy is current or complete. Offsite that is actually onsite is the third: a “cloud” backup that is actually a sync service, or an “offsite” drive that sits in the same room as the NAS.

The fourth mistake is backing up everything indiscriminately. Your operating system and applications do not need to be backed up; they can be reinstalled. Your personal files, documents, photos, and irreplaceable data are what matter.

The fifth mistake is ignoring the 30-day rule for ransomware. Some ransomware lies dormant for weeks before activating. If your cloud backup only keeps 30 days of versions, a dormant infection could encrypt your backup before you even know you are infected. Choose a service that keeps at least 90 days of versions.

type: bar
title: Copies needed to survive each failure mode
x: Drive failure, Theft, Fire, Ransomware
Copies needed: 2, 2, 2, 3

The chart shows why 3-2-1 is the minimum, not the ideal. Drive failure, theft, and fire each require only two copies to survive. Ransomware requires three, because the third copy must be offsite and versioned to escape both the encryption and the dormant-infection trap.

The afternoon setup

Set up the NAS and configure it to back up your computers nightly. Install the cloud backup client and point it at your important folders. Test a restore from each. Put a quarterly reminder on your calendar to test again.

That is it. Three copies, two media, one offsite. The oldest and most reliable data protection strategy in computing, implemented in an afternoon, costing less than a single data recovery. The cosmic microwave background has survived for 13.8 billion years because it was written everywhere at once. Your photos deserve the same principle, even if they are considerably less old.

If you want a local-first workspace that keeps your files on your own machine, with nothing leaving it, take a look at OMMAIS Desktop. It is built on the same principle as the 3-2-1 rule: your data belongs to you, on your hardware, under your control.

How to make a 3-2-1 backup systemPlays from YouTube when you press play

3-2-1 backup rule explained: protect your data from disasterPlays from YouTube when you press play

---

By OMMAIS: LongCat 2.5 Preview Free using OpenCode’s cloud API