You back up your files. Good start. But six months later your backup drive is a graveyard of files named backup_final, backup_final2, backup_ACTUALLY_final, and you have no idea which one still has the client contract from March - or whether the drive is even big enough to take a new copy before it fills up.
That's the problem a backup rotation scheme solves. It's not a fancy concept - it's just a rule for deciding which old backup gets replaced by a new one, and which ones you keep around longer. Get the rule right once, and you stop thinking about it. Get it wrong, and you either run out of space or realize too late that the one backup you actually needed got overwritten last Tuesday.
What a rotation scheme actually decides
Every rotation scheme answers the same two questions:
- How many backup copies (or versions) do you keep at once?
- When a new backup is due and you're out of room, which old one gets replaced?
That's it. The differences between schemes come down to how smart they are about which copy to sacrifice. A basic scheme throws away whatever's oldest, no questions asked. A more thoughtful one keeps a few old copies around on purpose - because "oldest" isn't always the same as "least useful."
This matters more than it sounds like it should. If ransomware or a bad update corrupts your data and nobody notices for two weeks, a scheme that only ever kept the last few days of backups won't save you. One that kept a copy from further back will - and that gap is exactly what recent industry research keeps flagging as the weak point for most organizations (more on that in the sources section below).
Comparing the main schemes at a glance
Here's the short version before we get into each one individually. None of these is "the correct" scheme - they trade storage cost against how far back you can reach.
| FIFO | Grandfather-Father-Son | Tower of Hanoi | 3-2-1 Rule | |
| What it decides | Which copy to overwrite | Which copy to overwrite | Which copy to overwrite | Where copies live |
| Rule of thumb | Newest replaces oldest | Daily / weekly / monthly layers | Doubling-interval sets | 3 copies, 2 media, 1 offsite |
| Storage cost | Lowest | Moderate | Moderate, more efficient over time | Depends on scheme paired with it |
| Restore depth | Shallow - a fixed window | Deep - weeks to months back | Deepest for the media used | N/A - survives site loss |
| Best for | Working files, scratch data | Most small businesses | Storage-conscious deep history | Everyone, alongside any scheme above |
FIFO - first in, first out
The simplest possible rule: the newest backup replaces the oldest one, full stop. If you keep 10 versions, the 11th backup overwrites the 1st. It's easy to set up and cheap on storage, which makes it fine for data that changes constantly and isn't mission-critical - a working folder, a scratch drive, daily project files. The catch is exactly the ransomware scenario above: FIFO has no memory beyond its window. We've got a full walkthrough of setting up FIFO backup rotation if that's the fit for your data.
Grandfather-Father-Son (GFS)
The old workhorse, and still the most common answer to "what's a sane backup schedule for a small business." It layers three frequencies - daily (son), weekly (father), monthly (grandfather) - so you keep a deep daily history for the last few weeks, a longer weekly history for a couple of months, and monthly snapshots stretching back even further. It costs more storage than FIFO, but it's the difference between "I can restore yesterday's version" and "I can restore the version from before things went wrong two months ago." Most small businesses that outgrow FIFO land here. Details and setup steps are on our Grandfather-Father-Son backup page.
Tower of Hanoi
Named after the puzzle, and it earns the name - instead of fixed daily/weekly/monthly slots, each backup set is used on a doubling interval (every 2nd day, every 4th, every 8th, and so on). The payoff is a much deeper history using fewer total copies than GFS would need for the same span. The tradeoff is that it's less intuitive to reason about at a glance - "which set holds last month's data?" isn't a one-second answer anymore. It suits people who are comfortable trading a bit of simplicity for storage efficiency. We break down the logic on the Tower of Hanoi backup strategy page.
The 3-2-1 rule
Technically this is about where your copies live rather than when they rotate, but it shows up in every rotation conversation for good reason: three copies of your data, on two different types of storage, with one of them offsite. Pair 3-2-1 with any of the schemes above - it tells you where the copies go, GFS or Hanoi tells you how long each one sticks around. If you haven't set this up yet, it's worth doing before you fine-tune anything else; see our 3-2-1 backup rule guide.
Worth knowing: a growing number of security researchers now recommend going one step further - 3-2-1-1-0, which adds one immutable or offline copy that ransomware physically can't reach, plus routine recovery testing. See the research note in the sources section below.
Tape rotation: still a thing, just not the default anymore
Rotation schemes were invented for tape. Back when a "backup" meant physically labeling a cassette and swapping it into a drive, someone had to decide which tape got reused this week - that's literally where Tower of Hanoi and GFS come from. If you've seen the terms "tape set," "media rotation," or "backup generation" used almost interchangeably with these schemes, that's the tape legacy showing through.
Do small businesses still use tape today? Some do, mostly where a physical, disconnected copy matters for compliance or insurance reasons - a drive that's unplugged and sitting in a fireproof safe can't be touched by ransomware, full stop. But for most home offices and small businesses, tape has been replaced by external drives, NAS boxes, and cloud storage that do the same job - swap in, swap out, keep a few generations - without needing a tape drive or a labeling system.
The good news: the scheme logic carries over even when the media doesn't. Handy Backup isn't built to connect to a physical tape drive directly — like most modern backup tools, it's designed around NAS, external drives, FTP/SFTP, and cloud storage instead. If you're moving off tape, GFS or FIFO logic applies exactly the same way to any of those; see how Handy Backup replaces LTO tape backup with modern storage options.
Picking a schedule without overthinking it
A few honest rules of thumb, based on what actually goes wrong for people:
- If you'd only be mildly annoyed to lose the last few days of a file, FIFO with a week or two of depth is enough.
- If losing that data would mean redoing real work or explaining it to a client, use GFS - the storage cost is small compared to the alternative.
- Whatever scheme you pick, at least one copy needs to leave the building (or the building's network). Tape, an external drive in a bag, or cloud - doesn't matter which, but "everything lives on the same desk" isn't a backup rotation, it's a single point of failure with extra steps.
If you want the practical, step-by-step version of turning any of this into an actual schedule - how often, what time of day, what to do around weekends and holidays - that's covered separately in our backup schedule best practices guide.
The scheme you pick matters less than picking one and automating it. The businesses that get burned aren't the ones running FIFO instead of GFS - they're the ones with no rule at all, relying on someone remembering to swap a drive. Whichever scheme fits your data, Handy Backup's scheduler will run it in the background so "remembering" stops being part of the plan.
Download for Free