If you’re regular about your file backups, you’re already miles ahead of most of my friends and relatives, so well done. But are you sure you’re not trusting those backups a little too much?
All too often, I’d seen someone with a regular backup schedule get too complacent and end up with data loss. Backup failures really sneak up on you sometimes.
“Backup complete” isn’t proof that you’re in the clear
It’s usually fine, but better safe than sorry.
Backups range from “I copied that one file elsewhere to keep it safe” to having a full-blown backup center with a NAS (or several), cloud backups, and off-site copies. No matter where yours lands on that spectrum, the same universal truth applies: Your backup may or may not be complete.
Most consumer backups are file-level. That means they copy what you point them at, basically. It could be a drive, a NAS share, or a cloud folder. The catch is that file-level backups can only back up what the system will reliably hand over.
That’s exactly why the most dangerous backup failures are the ones that are easy to miss. It’s not the stuff that fails; that, you’ll always notice. It’s the backups that seemingly complete without a hitch, but hide errors underneath the surface.
What your backup may have skipped
You won’t know until you need it.
Assuming you’re still inclined to trust your backups, let’s talk about the types of things that may get skipped, and whether this should worry you or not.
On Windows, open or actively changing files are a classic pain point for backups. To combat this, many backup tools rely on VSS snapshots to get a consistent point-in-time view of your data. If that snapshot step fails or is out of sync with the actual state of your data, you may end up with missing or inconsistent backups of the exact files you were working on.
Permissions can also be a problem. If the backup process hits files that it can’t read due to a lack of permissions, well, tough luck, that part of the backup isn’t happening. Some tools, like Robocopy, have a backup mode designed to override ACL blocks. If your backup runs without the right access, though, and the tool can’t work around that, “backup complete” will be more like “everything the tool could reach.”
Cloud folders used as backup targets or sync may feel like they’re safer, but they aren’t always. As an example, OneDrive’s Files On-Demand can mark some files as “online only,” which Microsoft explicitly describes as not taking up space on your PC and not downloading until you open them. A local backup job that’s copying what exists on disk may not capture what is not actually stored locally.
Meanwhile, some missing data is by design. Time Machine, for example, lets you exclude certain items from backups. This is useful and all, but I’ve definitely omitted things I shouldn’t have just by forgetting about excluding them in the first place.
When files copy just fine, but your backup still fails
As if backing up wasn’t annoying enough.
We’ve talked about sneaky backups that claim they’re fine when they’re not. Now, let’s talk about the ones that complete and look great at a glance, and there aren’t any missing files, and yet, something is still wrong.
I know, I know. What a nuisance.
This is the important distinction between just copying files and making sure they’re actually usable. Let’s say you copied a folder or two, and all the files safely made it across. File sizes look correct, too. But when you try to run one of the files, it won’t open.
That’s silent data corruption, and it can happen for any variety of reasons.
There could be something wrong with the source drive or the destination; the PC might have issues with the RAM or the CPU; perhaps one of the accessories is failing, such as a USB cable used to transfer files with. Data can be corrupted at any point on its way from point A to point B.
A big reason is that some data is not meant to be backed up as simple files while it’s live. Stuff like browser profiles, email stores, and anything that behaves like a database. You can copy those things, but if they were being written to at the time, you may find them corrupted on the target drive. At best, they’ll be missing recent changes. The same rule applies to anything with multiple moving parts, like a project folder. Access issues are common for those.
Another common trap is if you assume that restoring files is the same thing as restoring how those files worked on your old system. Permissions and ownership can come back different, especially if you’re dealing with a new device or new user account (and transferring across different OSes makes that trickier).
In the end, a backup can still succeed while also being the wrong backup, with missing files and errors that only pop up when you investigate each file.
How to test your backups (and keep them safe)
I’m the last person to advise you against backups. SSDs can fail for any reason, at any time, and the same can be said of any storage device. Backing up is the smart thing to do, and following the 3-2-1 rule (or, better yet, 3-2-1-1-0) is a must if you want to make sure your files are safe.
The key lies in never trusting your backups completely, and if the files are really important, going the extra mile toward securing them.
Start by testing your backups. Pick a small folder of important files, restore it to a new location (not on top of the original, but somewhere else entirely), and open what you restored one by one. It’s a chore, yes, but it’s worth the test. Make sure to include at least one archive (zip file, one large media file, and one file that’s been edited many times, because those may be more susceptible to data corruption.
If you want to check data integrity without the painful ordeal of going through it all file by file, PowerShell on Windows makes it easy. Just run the Get-FileHash command on the original and the restored copy and make sure the values are aligned. If not, that’s your first sign that something’s broken along your backup chain.
Aside from doing backups on top of backups, which is always a good idea, consider adding versioning and separation. Keep at least one backup target that supports older versions; that way, a mistake won’t overwrite your only copy. Then, keep one copy that isn’t always connected or always syncing to avoid having it get overwritten. Those steps will keep your files safer from disaster.
