IT Procurement, Maintenance & Troubleshooting

Backup and Disaster Recovery: A Practical Guide for Small and Medium Businesses

Disks fail, laptops are stolen, files are deleted and ransomware encrypts what it can reach. With good backups a business loses a few hours; without them it can lose years of records. This guide covers what to back up, how to protect the copies and how to plan the recovery.

  • Beginner
  • About 10 min
  • Last reviewed
Start the guide

Overview

What you will learn

  • The difference between a backup and a disaster recovery plan
  • How the 3-2-1 rule works: server, local NAS and off-site cloud
  • Why one copy must be offline or unchangeable
  • How to decide how often to back up (RPO) and how fast to recover (RTO)
  • How to test restores and write a short recovery plan

Applies to

  • Office servers and computers
  • NAS backup devices and external drives
  • Cloud backup services
  • Cloud email and file services such as shared drives

Tools Required

No tools needed.

Before you start

  • A list of the data and systems the business cannot work without

Important

  • A backup drive left connected to an everyday computer can be encrypted by ransomware along with everything else.
  • Never test a restore by overwriting live data. Restore to a separate test location.

Step-by-Step Guide

Building a 3-2-1 backup, step by step

1 / 8 steps

Office server Accounts PC Manager laptop Cloud email

Step 1

Copy one: the original data

  1. Your data lives on the office server, staff computers and cloud services. This is copy one, and it is the copy most likely to be lost.

    Expected result: You know where every important system's data lives.

  2. Back up the server and key computers to a NAS in another room. Local copies restore quickly after a deleted file or failed disk.

    Expected result: Fast restores for everyday mistakes and failures.

  3. Send an encrypted copy to a cloud backup service or another building. If the office suffers fire, flood or theft, this copy survives.

    Expected result: Three copies, on two devices, with one off site.

  4. Rotate a drive that is disconnected after each backup, or use backup storage that cannot be changed or deleted from the office network.

    Expected result: An attack cannot encrypt every backup at once.

  5. Put the NAS and network cabinet on a UPS so a power cut does not corrupt a backup in progress. Use separate backup accounts with multi-factor authentication.

    Expected result: Backups finish cleanly through short power cuts.

  6. Read the backup report each week. Every important system should show a recent successful backup, with failures followed up the same day.

    Expected result: Failed or missing backups are noticed before they are needed.

  7. Restore a file, a mailbox or a whole system to a test location. Note how long it took, including a restore from the off-site copy.

    Expected result: You know the backups work and how long recovery really takes.

  8. List who to call, the order systems come back, where backups and passwords are kept, and what staff do meanwhile. Review it after each test.

    Expected result: Anyone can start the recovery, not just one person.

Troubleshooting

Common problems and how to fix them

1 Backups report success but files are missing

Likely cause The backup job does not include every folder or system.

Solution Compare the backup selection with your list of critical data and run a test restore.

2 Backup drive encrypted along with the server

Likely cause The drive was permanently connected or reachable from the network.

Solution Rotate disconnected drives or use immutable backup storage, with separate backup accounts.

3 Off-site backup never finishes

Likely cause Too much data for the internet connection, or uploads stopped by power cuts.

Solution Seed the first copy locally, back up changes only, schedule uploads overnight and put the router on a UPS.

4 Restore takes days, not hours

Likely cause Recovery was never tested or the off-site link is slow.

Solution Time a full restore, keep a local copy for fast recovery, and plan replacement hardware in advance.

5 Deleted cloud files cannot be recovered

Likely cause The service's version history or recycle bin period has passed.

Solution Add a separate backup of important cloud data with longer retention.

6 Only one person knows how to restore

Likely cause No written plan or shared access.

Solution Write the one-page plan, store credentials securely, and have a second person run a test restore.

In Depth

In Depth

Disks fail, laptops are stolen, files are deleted and ransomware encrypts what it can reach. With good backups a business loses a few hours; without them it can lose years of records. This guide covers what to back up, how to protect the copies and how to plan the recovery.

Note Extrion provides backup and disaster recovery services and earns from them. No vendor paid for or reviewed this guide.

Backup and disaster recovery are not the same

  • Backup is a copy of your data that you can restore.
  • Disaster recovery is the plan for getting the business working again: which systems come back first, on what equipment, by whom and how long it may take.

You need both. A backup that nobody knows how to restore is little help on the day.

List what matters

Write down the data and systems the business cannot work without, for example:

  • Accounting and payroll data
  • Customer, patient or guest records
  • Business software databases
  • Shared files and documents
  • Email and calendars
  • Website files and database
  • Configuration of servers, firewalls, switches and CCTV recorders

For each one, note who owns it and where it lives today: a laptop, a server in the office or a cloud service. Anything that lives only on one laptop is your biggest risk.

Decide how much you can afford to lose

Two questions guide the design. IT people call the answers RPO and RTO:

  • How much work could you redo? (recovery point objective, RPO) If losing a day of entries would be painful, back up at least daily. If an hour would be painful, back up more often or continuously.
  • How long can you be without it? (recovery time objective, RTO) If a system must be back within hours, plan the equipment and steps in advance rather than buying a replacement after the failure.

Different systems get different answers. Accounts may need daily backups and a one-day recovery; a sales database may need hourly copies and a recovery within the morning.

Follow the 3-2-1 rule, and keep one copy offline

The UK National Cyber Security Centre describes the 3-2-1 rule as at least three copies of your data, on two different devices, with one off site. It adds that 3-2-1 on its own does not require any copy to be offline, so it recommends keeping at least one backup offline at any time: then an incident cannot affect all your backups at once (NCSC: Offline backups in an online world).

In practice, that often means:

  • The original, on the server, computers or cloud service.
  • A local backup device, such as a NAS in another room, for fast restores.
  • An off-site copy, in a cloud backup service or another building, in case the office suffers fire, flood or theft.
  • One copy ransomware cannot change: a disconnected drive that is rotated, or backup storage that keeps earlier versions and cannot be overwritten or deleted from the office network.

Protect the backups themselves

  • Use separate accounts for the backup system, with strong passwords and multi-factor authentication. Ransomware often uses stolen administrator accounts to delete backups first.
  • Do not leave backup drives permanently connected to everyday computers.
  • Encrypt backups that leave the building, and store the encryption key safely somewhere else.
  • Keep the backup device and the network cabinet on a UPS so a power cut does not corrupt a backup in progress. Frequent power cuts make this essential.

Test restores

A backup is only proven when you restore from it, and NCSC advises testing backups regularly. Each month, pick a file, a mailbox or a whole system and restore it to a test location. Write down how long it took and what went wrong. Restoring over a slow internet link can take far longer than people expect, so time an off-site restore at least once.

Cloud services still need backups

Files synchronised to a cloud drive are convenient, but a deleted or encrypted file is synchronised everywhere. Check how long the service keeps deleted files and earlier versions, and consider a separate backup of important cloud data such as email and shared drives. Our guide to cloud vs local hosting covers this choice in more detail.

Note If an off-site backup holds customer, patient or staff records and is stored outside Tanzania, the Personal Data Protection Act, 2022 rules on transfers abroad may apply, and the PDPC asks organisations to apply for a transfer permit first. Ask a legal adviser if you are unsure. This is not legal advice.

Write a short recovery plan

Keep it short enough that people will read it:

  1. Who to call, with phone numbers that do not depend on the office systems.
  2. The order in which systems are restored.
  3. Where the backups are and how to reach them, with passwords stored securely away from the systems themselves.
  4. What staff do in the meantime, such as paper forms or a temporary workspace.
  5. When the plan was last tested, and by whom.

NIST's Contingency Planning Guide for Federal Information Systems (SP 800-34 Rev. 1) is a detailed public reference if you want to go further. For protecting systems before anything goes wrong, try our cybersecurity checklist.

When to call Extrion

If you want backups set up, checked or tested, see our backup and disaster recovery service, our managed IT support, or contact us.

This guide is general information. Procedures, product features and firmware change, so check your own equipment's documentation before you start.

FAQs

Frequently asked questions

Is an external hard drive enough?

It is better than nothing, but one drive is one copy in one place. If it stays connected, ransomware can reach it; if it stays in the office, a fire or theft can take it with the computer.

How often should we back up?

As often as the amount of work you could afford to redo. Many businesses back up important data at least daily and keep several earlier versions, so they can go back to before a problem started.

Does RAID count as a backup?

No. RAID protects against one disk failing, but deleted, corrupted or encrypted files are lost on every disk at the same time.

Is OneDrive, Google Drive or Dropbox a backup?

Not on its own. These services synchronise changes, including deletions and encryption, to every device. Version history helps for a while, but a separate backup protects you from account loss, long-unnoticed problems and mistakes.

How long should we keep backups?

Long enough to go back before problems that take time to notice, such as a slow ransomware infection or a corrupted file nobody opened for weeks. Keep daily copies for a few weeks and monthly copies for longer, and check any legal or tax record-keeping rules that apply to you.

What should we do first if ransomware hits?

Disconnect affected computers from the network, do not pay or delete anything, and call your IT support. Check that the offline backup is safe before you connect it to anything.

Is Your Data Backed Up Properly?

We set up backups you can restore from, test them, and write the recovery plan with you.

Still stuck, or want more detail?

Ask for clarification, tell us a problem this did not solve, or suggest a topic. An engineer reads every message, and we use them to improve the Knowledge Center.

What would you like to do?
online now People browsing Extrion Tech right now

Scan to open this page

Point your phone camera at the code to open this page on your phone, or share it with someone nearby.

QR code that opens this page

https://mail.extrion.co.tz/knowledge-center/backup-and-disaster-recovery-guide

Download QR

Let’s start with you Who should our team speak to about this service?
How should we reach you?

Spam protection