Missing NAS Capacity: Where Did My Space Go?

A reader emailed me last spring in a mild panic. He had just built a Synology box with four brand new 4TB drives, twelve terabytes of usable space on paper after one drive of parity, and the dashboard was stubbornly reporting something closer to 10.5TB. He was convinced one of the drives had shipped half dead.

Nothing was dead. Every single gigabyte was accounted for, exactly where math says it should be.

This is the most common false alarm I see from new NAS owners, and it comes from a stack of small, boring subtractions that pile up into one alarming number. Let me walk through the questions people actually ask when this happens, in the order they usually ask them.

Why does my 4TB drive only show about 3.6TB?

This one trips up almost everyone, and it has nothing to do with your NAS at all. Drive manufacturers sell capacity in decimal terabytes, where one TB equals one trillion bytes. Your operating system, and most NAS dashboards, measure in binary units where the same physical space is divided by 1024 at each step instead of 1000.

So a 4TB drive is genuinely four trillion bytes. Divide that by 1024 three times and you get roughly 3.64 TiB, the binary tebibyte. Same bytes, different ruler.

If you want the deep version, the tebibyte explainer on Wikipedia lays out the 1024 versus 1000 history cleanly. The short version: you lose about 9 percent of the printed number to this unit mismatch alone, and you never actually had those bytes to begin with.

Where does RAID take my space?

Once the unit confusion clears, the next chunk of missing space is redundancy, and this one is real. RAID does not give you every drive added together. Some capacity is reserved to survive a drive failure.

The amount depends on the RAID level you chose. Here is how the common ones break down for a four bay NAS filled with 4TB drives, using decimal TB for the raw math.

That 12TB figure for RAID 5 is the raw decimal number. Run it back through the TiB conversion and you land near 10.9TiB, which is exactly the 10.5-ish figure my panicked reader was staring at. Parity plus unit math, stacked.

Does mixing drive sizes waste more?

Yes, with classic RAID it can waste a lot. Standard RAID 5 sizes every disk to match the smallest one, so pairing a 4TB drive with three 8TB drives throws away half of each big disk. Synology Hybrid RAID (SHR) is smarter and reclaims most of that mismatched space, which is one genuine reason to prefer it on a mixed pool.

What is filesystem and formatting overhead?

After parity, the filesystem itself skims a bit off the top. Btrfs and ext4 both reserve space for metadata, inodes, and internal bookkeeping, and many setups keep a small reserved block (often around 5 percent on ext4) so the system can still function when the volume is nearly full.

On a healthy NAS this is usually 1 to 3 percent of the volume, not a huge slice. But it is real, and it is why a freshly formatted empty volume never reports 100 percent free.

Swap and system partitions eat a little too. Synology and QNAP both carve out a few gigabytes across every drive for the operating system, which is why even a single-drive unit shows slightly less than the bare disk.

Why does my free space keep dropping when I am not adding files?

This is the sneaky one, and it is almost always snapshots. If you enabled snapshot schedules, or you run Btrfs shared folders with versioning, the NAS keeps point-in-time copies of your data. Every change you make holds onto the old blocks until the snapshot expires.

I once watched a client's 8TB volume quietly fill by 40GB a day. He swore he was deleting more than he added. He was, but every deletion was being preserved by an hourly snapshot policy nobody had reviewed in a year.

Recycle bins are the same trap on a smaller scale. If your shared folders have the network recycle bin turned on, deleted files sit in #recycle consuming full space until you empty it. On a busy media share that alone can hide hundreds of gigabytes.

How do I find what is actually using the space?

Stop guessing and measure. Every mainstream NAS ships a storage analyzer that scans your volume and ranks folders and file types by size. It is the single fastest way to find the real culprit.

1. Open your storage analyzer (Storage Analyzer on Synology, or the Storage & Snapshots panel on QNAP) and run a full report on the volume.
2. Sort shared folders by size and note the top three. Media and backup folders almost always dominate.
3. Check the snapshot total for each folder separately. This number is often invisible in the normal file view.
4. Empty every recycle bin, then re-check. Deleted files still counting against you show up here.
5. Look for orphaned backup versions from Time Machine or Hyper Backup, which love to accumulate silently.

Run that once and the mystery usually resolves in ten minutes. In my experience the ranking is nearly always snapshots first, forgotten backups second, and duplicate media third.

Is any of my space actually lost or wasted?

Rarely, but occasionally yes, and it is worth ruling out. A drive that failed a SMART test and got dropped from the pool can leave a volume running in degraded mode with a bay's worth of capacity offline. Check your storage manager for any drive flagged as warning or critical.

The other genuine waste is thin-provisioning gone wrong on business-oriented units, or a volume you created smaller than the underlying storage pool. If your pool shows more free space than your volumes combined, you may simply have room you never allocated. Expanding the volume reclaims it instantly.

For the difference between how vendors count all this, the SNIA storage dictionary is the reference the industry itself uses, and it settles most terminology arguments.

Putting the numbers back together

Here is the honest accounting for that four-bay, four 4TB drive, RAID 5 build that started this whole thing. Raw hardware is 16 decimal TB. One drive of parity removes 4TB, leaving 12TB usable. Convert to the binary units the dashboard uses and it becomes about 10.9TiB. Trim a percent or two for filesystem overhead and system partitions, and you settle around 10.6TiB free on day one.

Not a single byte went missing. It went to redundancy, to honest math, and to the small tax every filesystem charges.

Related reading: NAS Hard Drive Failure: Signs, Causes and Fixes, SMART Attributes That Predict Drive Failure and Pre-Replacement Checklist for a Dying NAS Drive.

The next time your NAS reports a number that looks wrong, resist the urge to assume a bad drive. Walk the same subtractions in order, unit conversion, parity, overhead, snapshots, recycle bins, and the gap almost always closes on its own. Once you have seen where it goes, you stop worrying and start planning your next capacity upgrade with numbers you actually trust.