Most small businesses we talk to have "a backup." Usually that means an external drive plugged into the server or the owner's PC, copying files every night. It is a good start, and it will save you if a single hard drive dies.
It will probably not save you from the threat that has been keeping IT people busy all year: ransomware. Programs like CryptoLocker and its successor, CryptoWall, encrypt your files and demand payment to unlock them, and they are very good at finding that backup drive sitting right next to your data. The 3-2-1 backup rule is the simplest way we know to make sure one bad day does not become a disaster.
The rule in one sentence
Keep three copies of your important data, on two different types of storage, with one copy kept offsite.
That is it. It is not tied to any product or brand. It is a way of thinking that protects against a wide range of problems, from a failed drive to a stolen laptop to a flooded office.
Three copies
Your working data counts as one copy. You need at least two more. The idea is simple probability: the chance of one copy failing is real, the chance of all three failing at the same time is very small, provided they are not all exposed to the same risk.
Two types of storage
Different storage fails in different ways. If your data lives on a server and your backup is on an external USB drive, a single power surge or a single piece of malware is less likely to take out both. Examples of different storage include a server's internal drives, a network backup appliance, external drives, tape, and cloud storage.
One copy offsite
This is the part most small businesses skip, and it is the most important one. If every copy is in the same building, a fire, theft, flood or lightning strike can take them all at once. Here in Tampa Bay, with hurricane season running through November, that is not a theoretical concern. An offsite copy can be in the cloud, at a secure data center, or on a drive that is rotated to another location.
If all your backups are in the same room as your server, you do not have a disaster recovery plan. You have a hardware failure plan.
Why ransomware changes the math
CryptoWall has been one of the most widespread threats of 2015, and the FBI issued a public warning about it this summer. It typically arrives through an email attachment or a compromised website. Once it runs, it quietly encrypts documents, spreadsheets, photos and databases, then displays a ransom note demanding payment in Bitcoin.
Here is the part that catches businesses off guard. Ransomware does not just encrypt the files on the infected PC. It encrypts anything that PC can reach, including mapped network drives and any external drive that happens to be plugged in. A backup drive that is always connected and always accessible looks, to ransomware, like just another drive full of files to lock up.
The defenses the 3-2-1 rule builds in are exactly what ransomware struggles with:
- An offsite or cloud copy that is not mapped as a drive letter on your computers cannot be reached by ransomware running on them.
- Versioned backups that keep multiple past copies let you restore files from before the infection, even if the most recent backup captured some encrypted files.
- A rotated drive that is disconnected most of the time is out of reach when an infection happens.
Sync is not backup
A common misunderstanding deserves its own section. File sync services like Dropbox, Google Drive and OneDrive are wonderful for sharing and working from anywhere, but syncing is not the same as backing up. When ransomware encrypts a synced folder, the encrypted versions sync right up to the cloud and replace the good ones. The same goes for accidental deletions. Some of these services keep previous versions for a limited time, which can help, but they should not be your only protection.
Test your restores
A backup you have never restored from is a hope, not a plan. Backups fail silently more often than anyone likes to admit: a drive fills up, a password changes, a job stops running after an update, or the software backs up the wrong folder. Everything looks fine until the day you need it.
Testing does not have to be elaborate:
- Monthly, pick a few random files and restore them to a different location. Open them and make sure they work.
- Quarterly, try restoring something bigger, like a whole folder or a database, and time how long it takes.
- At least once a year, walk through what it would take to rebuild an entire server from backup. The answer to "how long would we be down?" is worth knowing before a storm, not after.
Your 3-2-1 checklist
- List the data your business cannot live without: files, email, accounting, databases, line-of-business applications.
- Confirm you have at least three copies of each.
- Make sure those copies are on at least two different kinds of storage.
- Get one copy offsite, ideally with automatic cloud backup.
- Keep backup storage from being permanently mapped or accessible to everyday user accounts.
- Use backups that keep multiple versions going back weeks, not just last night.
- Check backup reports daily and test restores monthly.
- Write down who does what if you ever need to recover everything.
The 3-2-1 rule is not complicated, and following it turns a ransomware infection or a lost server from a business-ending event into an inconvenient afternoon. If you would like a second opinion on your current setup, our backup and disaster recovery services are designed around this approach. Reach out through our contact form.
