The 3-2-1 Backup Rule and a Disaster Recovery Plan That Actually Works

Backups are only useful if you can restore them quickly. Learn the 3-2-1-1-0 rule, how to set realistic recovery targets and how to write a disaster recovery plan your team can actually follow.

Opened laptop hard drive showing the platter
Backups and disaster recovery — Evan-Amos (CC BY-SA 3.0)

Almost every business says it has backups. Far fewer can say how long it would take to restore everything after a ransomware attack, a failed server or an accidentally deleted folder — or when they last actually tried. A backup you have never restored is a hope, not a plan.

This guide explains the modern 3-2-1-1-0 backup rule, how to set realistic recovery targets, what to back up (including cloud services people often forget) and how to write a disaster recovery plan that your team can follow under pressure.

Key takeaways

  • Define RPO (how much data you can lose) and RTO (how long you can be down) for each system before choosing tools.
  • Follow 3-2-1-1-0: three copies, two media, one off-site, one immutable or offline, zero restore errors.
  • SaaS data in Microsoft 365, Google Workspace and similar tools also needs backup.
  • Test restores on a schedule and write the results down.
  • Keep the disaster recovery plan short, printed and known to the people who will use it.

Backup is not the same as disaster recovery

Backup is a copy of data you can restore. Disaster recovery (DR) is the process and infrastructure that bring the business back to operation after a serious incident: restoring data, rebuilding systems, re-establishing access and communicating with staff and customers. Business continuity is wider still: how the business keeps operating at all while systems are down. You need backups to do DR, but backups alone do not guarantee a quick recovery.

Step 1: Set recovery targets — RPO and RTO

  • Recovery Point Objective (RPO): the maximum amount of data, measured in time, you can afford to lose. An RPO of 24 hours means a daily backup is acceptable; an RPO of 15 minutes requires much more frequent copies or replication.
  • Recovery Time Objective (RTO): the maximum time a system can be unavailable before the impact becomes unacceptable.

Set these per system, together with the people who use them, and be honest about cost: shorter RPO and RTO cost more.

SystemExample RPOExample RTOTypical approach
Accounting / ERP database1 hour4 hoursFrequent database backups + documented rebuild
Email & documents (SaaS)24 hours8 hoursThird-party SaaS backup with granular restore
E-commerce website1 hour2 hoursDatabase + files backup, infrastructure-as-code
File archive24 hours2 daysDaily backup to cloud storage
Employee laptops24 hours1 dayData in cloud storage, device re-imaging

Step 2: Apply the 3-2-1-1-0 rule

The classic 3-2-1 rule has been extended to address ransomware:

  • 3 copies of your data (the original plus two backups)
  • 2 different types of storage (for example a local appliance and cloud object storage)
  • 1 copy off-site, in a different location or region
  • 1 copy immutable or offline — it cannot be changed or deleted, even by an administrator account, for a set retention period
  • 0 errors in your most recent restore test

The immutable copy is the one that saves you when attackers gain administrator access and try to delete backups before encrypting your systems. Most modern backup tools and cloud storage services offer object lock or immutability features.

Step 3: Know what to back up

  • Databases of line-of-business applications, using application-consistent methods.
  • File shares and document libraries.
  • SaaS data: email, OneDrive/SharePoint, Google Drive, CRM, accounting exports. Providers keep their service running, but under the shared responsibility model they generally do not protect you against deletion by a user, a malicious insider or a compromised account beyond limited retention windows.
  • Websites: database, uploaded media, themes and plugins or application code.
  • Configuration: firewall and router settings, DNS records, cloud infrastructure definitions, licence keys.
  • Credentials and documentation needed for recovery, stored securely and accessible during an outage (for example in a password manager with emergency access).

Step 4: Choose retention

A common scheme keeps daily backups for 30 days, weekly backups for three months and monthly backups for one year. Adjust for legal retention requirements and for the fact that some attacks or data errors are discovered weeks later. Remember that backups containing personal data are subject to the same privacy obligations as the original data.

Step 5: Protect the backups themselves

  • Use separate credentials for the backup system, protected by MFA, not linked to your everyday admin accounts.
  • Encrypt backups at rest and in transit, and store the keys safely.
  • Enable immutability or keep an offline copy.
  • Monitor backup jobs and alert on failures — a silent failure for three months is common.

Step 6: Test restores regularly

Schedule restore tests and record the results:

  • Monthly: restore a few random files and a mailbox item.
  • Quarterly: restore a complete database or virtual machine into an isolated environment and verify the application works.
  • Annually: run a tabletop or full DR exercise based on a realistic scenario, such as ransomware on the main server.

Measure how long each restore takes and compare it with your RTO. If restoring 2 TB from cloud storage takes 30 hours and your RTO is 8, you need a different approach — for example a local copy for fast restores plus the cloud copy for disasters.

Step 7: Write the disaster recovery plan

A good DR plan is short enough to be used under stress. Include:

  1. Roles and contacts: who leads, who decides, IT partner, key vendors, insurer, bank, legal advisor — with phone numbers.
  2. Triggers: what counts as a disaster and who declares it.
  3. Priorities: the order in which systems are restored, based on RTO.
  4. Procedures: step-by-step restore instructions for each critical system, with where backups and credentials are.
  5. Communication: templates for staff, customers and suppliers, and an alternative channel if email is down.
  6. Workarounds: how the business keeps operating manually while systems are restored.
  7. After-action review: how you will learn and improve after an incident.

Store the plan in several places, including on paper. If your systems are encrypted, a plan stored only on the file server is useless.

Ransomware-specific considerations

  • Isolate infected systems quickly, but preserve evidence.
  • Restore into a clean environment; attackers often remain in the network for a while before encrypting, so check backups for persistence mechanisms.
  • Reset all credentials, starting with administrators, from clean devices.
  • Check legal notification duties if personal data may have been accessed.
  • Prevention matters most: see our cybersecurity checklist.

A simple DR plan template

1. Incident lead: name, phone · Deputy: name, phone
2. IT partner: company, 24/7 number · Insurer: policy no., hotline
3. Restore priority: (1) identity & email, (2) accounting, (3) website/shop, (4) file shares
4. Backup locations: local appliance (office), cloud storage (region), immutable copy (retention 30 days)
5. Credentials: password manager emergency access held by lead and deputy
6. Communication: staff via phone tree / messaging group; customers via website banner and template email
7. Last full test: date, duration, issues found, actions

Frequently asked questions

Is cloud sync (OneDrive, Dropbox, Google Drive) a backup?

Not on its own. Sync replicates deletions and encrypted files too. Version history helps, but a separate backup with independent retention is still recommended.

How often should we back up?

As often as your RPO requires. Daily is a common minimum; critical databases often need hourly or continuous protection.

How long does a restore take?

It depends on data volume, bandwidth and storage type. Measure it in a test — that number is far more useful than any estimate.

Do we need a DR plan if everything is in the cloud?

Yes. Accounts can be compromised, data can be deleted and providers can have outages. The plan may be simpler, but it is still needed.

Need help?

Backup monitoring and quarterly restore tests are part of our Managed IT Support plans, and our Cloud Migration & DevOps assessment includes a backup and DR design with defined RPO and RTO. Contact us if you would like someone to check whether your backups would really save you.

Sisusoft Team

Practical guides from the engineers and IT specialists at Sisusoft Technologies. We write about the problems we solve every day for small and growing businesses.

Keep reading

Related articles

Have a problem that technology should solve?

Tell us about it. We reply within one business day with honest next steps.