What Snapshots Are and How They Differ from Backups

Snapshots and backups are two terms that, in conversations about servers and data storage, often get blurred into one. In reality, the difference between them is like the difference between saving a video game and evacuating valuables to a bunker. Their overall goal is similar — preserving information. But the methods, speed, and reliability are fundamentally different.

A Backup Is Standalone Insurance

Imagine a mansion filled with furniture, archives, and fragile items, owned by someone who is deeply concerned about security.

A backup, in this analogy, is a truck that arrives once a day, carefully packs up every valuable item, and hauls it off to a secure warehouse on the other side of town. If the mansion burns to the ground or gets robbed, everything can be restored from scratch — just move the items back from the warehouse.

In the digital world, a backup works exactly the same way. It is a self-contained set of files holding an exact copy of the data as it existed at the time of creation. It is stored on a separate physical medium, in a different data center, or in the cloud.

Creating a backup is not a fast process, but it is completely autonomous. A server failure, a lightning strike, or a ransomware attack cannot touch it. Even if the original is destroyed, the copy remains intact. The price of that reliability is time, disk space, and the fact that restoring from a backup is rarely instantaneous — the data has to be moved to a new location.

A Snapshot Is an Instant Impression of a State

Continuing the mansion analogy — imagine a risky rearrangement is planned. The problem is that all the furniture is extremely heavy, and moving it risks scratching the floors or damaging the walls. So it makes sense to take an ultra-precise three-dimensional photograph of the room, capturing the exact position of every object. Then you can experiment with the layout freely, without fear of causing any damage.

A snapshot is exactly that three-dimensional photograph. It is created almost instantly and takes up almost no space, because it does not copy the data itself — it records references to it. From that point, the system operates on a delta basis: if data begins to change, the original blocks are preserved in a reserved area while new data is written separately. As long as the snapshot is active, the user sees a coherent picture, but no physical duplication of data occurs. If the rearrangement turns out to be a disaster, rolling back to the snapshot restores everything to its original state in seconds.

But that ease comes with a limitation. A snapshot lives on the same disk array as the original. It cannot be separated from its source. If the hardware itself fails, it takes both the data and the snapshots with it. This is a tactical tool for short time windows.

What Happens Inside: Pointers Instead of Copies

The technical mechanism behind a snapshot is elegant in its efficiency. Instead of rewriting an entire volume, the system freezes the current table of pointers to data blocks. When a command arrives to modify a block, the system leaves the original untouched, writes the changed data to free space, and updates the pointer. This preserves both the old and new versions while using minimal additional space.

A backup works more straightforwardly: it takes the source volume and fully duplicates it to another location. It does not conserve space, but it is maximally simple to restore from — the copy is ready to use with no dependencies on the original environment.

Why a Snapshot Is Not a Replacement for a Backup

In practice, beginners often fall into a trap. They configure frequent snapshot creation, appreciate the speed, and skip investing resources in traditional backups. Then the disk array fails, and all those “saves” vanish along with the main system. The snapshots were on the same hardware and shared its fate.

From this comes a simple rule: a backup must always be physically isolated from the primary storage. A snapshot, by contrast, can only exist within the same disk environment. One protects against catastrophes; the other protects against your own mistakes, failed updates, or short-term experiments.

Full Insurance vs. the Save Button

A backup is a comprehensive insurance policy. It requires regular premiums — time and space — but in the event of a true catastrophe, it fully restores what was lost, even if recovery takes a few hours or a day.

A snapshot is the “Save” button before a boss fight. Pressing it is nearly free, instantaneous, and lets you undo a bad move. But if the console or computer breaks, the save files are gone along with the game.

A sound data storage strategy does not pit these tools against each other — it combines them. Throughout the day you can freely experiment under the protection of snapshots, while at night a heavy, reliable backup quietly travels to a geographically remote storage location. Either one without the other holds up just fine — until the first serious incident. And it is better to remember that before it happens.