# Snapshot, backup and RAID: what differences for your data?

[Corresponding web page](https://wisphost.com/blog/snapshots-vs-backups/)

Rolling back before an update and restoring a lost server are two different needs. Examine the location of each copy, its dependencies and its restoration.

By WispHost. Published on 2026-11-01.

Topic: Infrastructure.

## The short answer

A snapshot represents a state at a given moment, a backup organizes data recovery and RAID concerns how disks are organized. Their protection depends on the implementation: verify that copies are independent, how long they are retained and the result of a restore test.

- A local snapshot may depend on the same disks as the server.
- RAID alone does not protect against deletion or an application error.
- A backup must be restored in a test to verify what it can recover.

## What is a snapshot for?

A snapshot preserves the representation of a state at a given moment. Depending on the technology, it may rely on block differences and continue to depend on the original data. It can make it easier to roll back before a change, but its name does not prove that an independent copy of the server exists.

Behaviors vary: VMware snapshots depend on their base disks, whereas AWS describes its EBS snapshots as separate backups. These examples explain why you must verify the implementation, without transferring its rules to another hosting provider.

## What should you expect from a backup?

A backup must make it possible to recover data after a defined event: human error, corruption, loss of the server or another incident covered by your plan. Verify the copy's destination, the access that allows it to be deleted and any dependency on the main environment.

Specify what is copied: files, database, configuration, disks or the entire machine. Copying files while the application is writing may require consistency precautions. Follow your software's procedure, particularly for exporting or restoring a database.

## Why does RAID not replace these copies?

RAID combines several disks according to a given organization. Some levels add redundancy against disk failure; others prioritize another objective. It does not automatically provide a history of your files or a copy outside the machine.

A deletion performed by the application may be propagated to the disks in the same set. You must therefore distinguish storage continuity from the ability to return to a previous version. The two functions can complement each other, but they do not replace each other.

| Mechanism | Essential question |
| --- | --- |
| Snapshot | What data and storage does this state depend on? |
| Backup | What can I restore if the server disappears? |
| RAID | What disk failures can this set tolerate? |

## Define acceptable loss and downtime

Choose how much data your business can lose between two copies and the target time to get the service running again. These are two different objectives. A daily backup may leave more changes to recover than a more frequent copy, depending on when it runs.

Also define how long versions are retained. An error discovered after several days requires a restore point that is still available. Copy frequency alone does not tell you how many versions you can view or how long a restore will take.

## How should you read the WispHost option?

The VPS catalogue distinguishes the base snapshots from the optional encrypted daily backups, with copies outside the host server. This option costs 3.50 USD per month before any duration discounts. It is reserved for VPS and does not apply to the dedicated or GPU ranges.

Before adopting a strategy, have the retention, covered content, access and restoration procedure actually provided clarified. NVMe storage or the RAID of another offering does not mean that this option would be included there.

## Test a restore before you need it

Restore a copy in a suitable test environment, then check the files, application startup and important data. Note the recovery point and the necessary steps, without exposing your private access. This gives you a procedure that your team can actually reproduce.

Repeat this after a major software or data change. A file created without errors is a useful indicator; above all, verify that you can recover a consistent service.

## Sources and references

- [Broadcom — dependencies and limitations of VMware snapshots](https://knowledge.broadcom.com/external/article/318825/best-practices-for-using-vmware-snapshots.html)
- [AWS — EBS snapshots used as backups](https://docs.aws.amazon.com/ebs/latest/userguide/ebs-snapshots.html)
- [IBM — RAID and the distinction from a backup solution](https://cloud.ibm.com/docs/bare-metal?locale=en&topic=bare-metal-bm-raid-levels)

## Your next step

[View the VPS and the backup option](https://wisphost.com/vps/)

## Read next

- [How many vCPU, RAM and how much storage should you plan for an VPS?](https://wisphost.com/blog/vps-resource-sizing/)
- [VPS or dedicated server: how to choose for your project?](https://wisphost.com/blog/vps-or-dedicated-server/)

[All WispHost guides](https://wisphost.com/blog/)