INFRASTRUCTURE & CLOUD
Tested backups: the difference between recovering and losing everything
Having a backup is not the same as being able to restore it. Learn the 3-2-1 rule, the concepts of RPO and RTO, why so many backups fail when you need them and how to set up a simple testing routine.
IN 30 SECONDS
- Follow the 3-2-1 rule, with at least one offline or immutable copy.
- Decide how much data the company can afford to lose (RPO) and how long it can be down (RTO).
- A backup that has never been restored is an assumption, not a guarantee.
- Cloud files and email need their own backup too.
Why so many backups fail
Many companies only find out their backup doesn’t work at the worst possible moment: after a ransomware attack, a disk failure or a file deleted by mistake. The reasons tend to repeat.
- The job stopped running and nobody noticed the alert.
- The backup was on the same network and got encrypted along with everything else.
- Only some of the systems were included.
- Restoring takes far longer than the company can afford to be down.
- Nobody knows the procedure, because it was never tested.
The 3-2-1 rule
- 3 copies
The original data plus two backup copies. - 2 different media
For example, local storage and the cloud, so a single failure can’t hit every copy. - 1 off-site
At least one copy outside the company and, ideally, offline or immutable, so it can’t be deleted over the network.
The next step: 3-2-1-1-0
Many teams use an extended version: one more offline or immutable copy and zero errors when verifying restores. The “0” is a reminder that a backup only counts once it has been tested.
RPO and RTO: the two questions that define your backup
| Concept | Question | Example |
|---|---|---|
| RPO (recovery point objective) | How much data can the company afford to lose? | A backup every 4 hours means losing up to 4 hours of work |
| RTO (recovery time objective) | How long can the company be down? | The ERP must be back within 2 hours |
These answers come from the business, not from IT. A sales system usually requires a short RPO and RTO; a historical archive can tolerate more. Once they are defined, it becomes clear how much to invest in backup and redundancy.
The cloud is not a backup
Services such as Microsoft 365 and Google Workspace keep the platform running, but protecting your data against deletion, human error or attacks is largely the customer’s responsibility. The recycle bin and version history only go back so far. Email, OneDrive, SharePoint and Google Drive need their own backup too.
A simple testing routine
| Frequency | Test |
|---|---|
| Daily | Check that backups ran and handle any alerts |
| Monthly | Restore randomly chosen files and folders |
| Quarterly | Restore a complete system in a test environment |
| Yearly | Simulate a disaster and measure the actual recovery time (RTO) |
From backup to business continuity
- List what is critical
Systems, databases, files and email the company can’t operate without. - Define RPO and RTO
Together with the business areas, for each critical item. - Implement the 3-2-1 rule
With an offline or immutable, encrypted copy. - Monitor and test
Alerts handled every day and regular restores recorded. - Document the procedure
Who does what, in what order and where the credentials are, so you don’t depend on a single person.
Quick backup checklist
- List of critical systems and data
- RPO and RTO defined with the business
- Three copies on two different media
- One offline or immutable copy
- Backup of cloud email and files
- Restore tested in the last 3 months
- Documented recovery procedure
HOW GESTÃO DE TI CAN HELP
IT Infrastructure & Networks
Backup, servers and networks planned and monitored so operations never stop.
For information only. This does not replace a professional review of your specific case. Legal references reflect the rules in force on the publication date (Brazilian law).