· managed it · 5 min read

Immutable Backups and the Restore Test Nobody Schedules

Most organizations have backups. Fewer have restored from them. This post covers 3-2-1-1-0, why attackers hit backups first, Microsoft 365 backup, and a restore testing schedule that internal IT and an MSP can share.

Most organizations have backups. Fewer have restored from them. This post covers 3-2-1-1-0, why attackers hit backups first, Microsoft 365 backup, and a restore testing schedule that internal IT and an MSP can share.

Picture an accounting firm in the second week of March. A server fails on a Tuesday morning, and forty people have client returns open in half-finished states. The IT manager says the backup ran last night.

Whether that helps depends on one thing: has anyone ever restored from it? A green checkmark on the backup dashboard means the job ran. It does not say the data comes back, in time, in a usable state. The only way to learn that is to restore something on a schedule and write down what happened.

3-2-1 and the newer 3-2-1-1-0

The old rule is 3-2-1: three copies of your data, on two kinds of storage, with one copy offsite. It has held up well for years.

The newer version adds two digits. The extra 1 is a copy that is offline or immutable. Immutable means nobody can change or delete it for a set period, including an administrator. The 0 means zero errors when you verify a restore.

That last digit is the one that gets skipped. A backup you have never restored is a guess.

In practice, the immutable copy is often a cloud repository with retention locking turned on, or a drive that sits disconnected between runs. Either works. What matters is that the credentials used day to day cannot reach it.

Why attackers go after backups first

Ransomware crews know that a working backup ends the negotiation. So they look for the backup console and the network shares before they encrypt anything. If your backups sit on the same network, behind the same admin password, they can be deleted or encrypted along with the servers.

An immutable or offline copy sits out of reach of those credentials. That is the copy you will want on the worst day.

Microsoft 365 needs its own backup

Many companies assume Microsoft 365 backs itself up. Microsoft keeps the service running, and it offers recycle bins and retention settings. Those are short windows, and they do not cover every mistake.

Picture an employee who deletes a shared SharePoint folder and empties the recycle bin. Or an attacker in one compromised account who wipes a mailbox. If you notice after the retention window closes, the data is gone.

A separate backup of mail, files, and Teams or SharePoint content gives you a way back. It needs the same restore testing as everything else.

RTO and RPO in plain words

Two terms come up in every backup conversation.

RTO (recovery time objective) asks how long the business can wait. If the system your staff use to take orders, bill clients, or answer calls is down, how many hours before you are turning customers away?

RPO (recovery point objective) asks how much work you can redo. If backups run nightly, you could lose up to a day of entries and edits. If that is too much, the backups need to run more often.

An accounting firm might accept a day of downtime on its document archive but only an hour on the tax software. Those are different answers, so they get different backup settings and different costs. Write both numbers down for your three or four most important systems.

A restore testing schedule

Here is a schedule that small teams can actually keep.

  • Monthly: restore a handful of files and one mailbox item, and have a real user confirm they open
  • Quarterly: restore a full server or a business database to an isolated spot, and time it
  • Once a year: walk through a full-site loss on paper, using the real recovery times from the quarterly tests
  • After any major change, such as a new server or a switch to a different backup product

Keep the test small enough that people do it. A fifteen-minute monthly check beats an annual project that keeps slipping.

Record the date, who ran it, how long it took, and whether it worked. That log is also what a cyber insurance carrier asks about on the renewal form.

Who runs restores in co-managed IT

In co-managed IT, restore testing is a good example of shared responsibility.

The MSP usually runs the backup platform, watches for failed jobs, keeps the immutable copy, and performs the heavy restores. Your internal IT team knows which data matters and what “correct” looks like. They pick the test files, and they ask someone who works in that system every day to open the restored data and confirm it.

Both sides sign the log. Neither side should be the only person who knows where the recovery steps are written down.

How we help

We run data backup and recovery with immutable copies, Microsoft 365 backup, and scheduled restore tests, and we hand your IT manager the results each time. Together we set the RTO and RPO for each system and test against them. Your IT manager keeps the authority over what matters most; we take the monitoring, the after-hours failures, and the heavy restores.

Call us at 405-253-5101 or schedule a discovery call to get on the calendar.

  • co-managed-it
  • backup
  • ransomware
Share:
Back to Blog