Guide · data recovery
APFS drive won't mount? What space manager invalid means, and how to get your files back
You copied files to an external drive, the cable got knocked, or the Mac went to sleep or the power dropped mid-copy, and now the drive won't mount. Disk Utility shows it greyed out. First Aid runs, churns, and fails, and macOS offers to erase it. Do not erase it. In most of these cases your files are still sitting on the disk, untouched. What broke is the drive's index of free space, the part APFS calls the space manager, not the files themselves. Here is what is actually happening, and how to get the data back without making it worse.
Last reviewed: 21 July 2026 · by Alien IT Solutions
What space manager invalid actually means
APFS keeps a map of which parts of the disk are used and which are free. That map is the space manager. When a write is cut off partway, by an unplugged cable, a crash or a flat battery, that map can be left half updated and inconsistent.
Run a read-only check and you will see it plainly:
sudo fsck_apfs -n /dev/diskNs1
On a drive like this, the top-level container checks out fine, then the check dies at the space manager with something like:
warning: (oid 0x6feb) cib: invalid o_cksum (0x0) error: failed to read spaceman cib 20 at address 0x6feb The volume /dev/diskNs1 could not be verified completely
The important part: the check never even reached your files. It tripped over the free-space map first. Your photos, documents and everything else live in a different structure, the file catalogue, that the interrupted write never touched. That is why this looks catastrophic but usually is not.
Why Apple's repair can't fix it
The obvious next move is to let macOS repair it:
sudo fsck_apfs -y /dev/diskNs1
It will not help. On space-manager corruption, Apple's repair tool bails out at the same spot and repairs nothing: the output is byte for byte identical to the read-only check. There is no built-in rebuild-the-free-space-map command in macOS. First Aid in Disk Utility is the same tool with a friendlier button, so it fails the same way.
Running First Aid over and over does nothing but risk more writes to a drive that is already in a fragile state. Stop doing it.
The two things not to do
Both of these feel productive. Both make it worse.
! Don't reformat or erase
The data is still there. Erasing throws it away. macOS will offer this the moment a volume fails to mount, and it is the one button you must not press while the files still matter.
! Don't keep mounting and replugging
Every mount, unplug and replug is another chance for a partial write. If the data matters, the drive should be touched as little as possible until it has been copied.
Step 1: clone the drive before you touch anything else
The golden rule of recovery: work on a copy, never the original. Make a full sector-by-sector image of the drive first, then do all the recovery work on that image. If anything goes wrong, the original is still exactly as it was.
On a Mac, or a Linux box, the tool for this is ddrescue, which reads around bad spots instead of giving up on them:
sudo ddrescue -r3 /dev/rdiskN /path/to/backup/drive.img /path/to/backup/drive.log
A clean 4 TB drive images in roughly thirteen hours and, if the hardware is healthy, comes back with zero read errors, which by itself tells you the problem is logical, a scrambled index, not a dying disk. The log mapfile means you can stop and resume without starting over.
You need somewhere to put the image: a second drive at least as big as the one you are rescuing. Never image a drive back onto itself, and never onto the drive you are rescuing from.
Step 2: recover from the clone with a tool that rebuilds the metadata
Because macOS cannot repair the space manager, you use recovery software that ignores it entirely and rebuilds the file list from the drive's other metadata. Tools such as UFS Explorer or R-Studio, named here as examples rather than endorsements, open the disk image read-only, scan it, reconstruct the APFS volume from what is intact, and can even fall back to a slightly older checkpoint if the newest one is the damaged one.
Open the image read-only
Open the disk image, not the physical drive, and keep it read-only the whole way through.
Scan the APFS volume
Point the tool at the APFS volume and run a full scan so it can rebuild the file list.
Preview before you commit
Most tools let you confirm files actually open before you commit, so you know the recovery is real.
Save to a third drive
Save recovered files to a third location with enough free space: a 3.9 TB volume needs a target of at least 4 TB. Never save back onto the image or the original.
If the previews open cleanly, you have your data. Spot-check a handful of the important files after saving.
When to stop and get help
DIY recovery is reasonable when the drive is healthy and the data is annoying to lose. Get a professional involved when any of these is true.
The drive makes noise or drops out
Clicking, buzzing, or disappearing and reappearing is physical failure, and every power-on makes it worse. Power it down and stop.
ddrescue is hitting read errors
Read errors that are not zero mean the surface is degrading. That is a different job to a scrambled index.
The volume is encrypted
If the volume is encrypted with FileVault and you don't have the password or key, the recovery changes shape. A lost key is its own problem; on Windows, a missing BitLocker recovery key blocks access even to a healthy drive.
The data is irreplaceable
If you would rather not gamble your one good clone on a first attempt, hand it over. A recovery service works from a clone exactly like the one above; they just have the tools and the hours to do it safely.
How to stop it happening again
Eject before you unplug. Safely remove is not nagging. It is what flushes the last writes and closes the free-space map cleanly.
Don't run a laptop copy on a dying battery. A shutdown mid-write is the classic cause of this exact fault.
Keep a second copy. A single external drive is not a backup. It is the thing that fails. One extra copy, a second drive or the cloud, turns a disaster into a mild annoyance. The same first-hour discipline applies to any storage scare: our dead drive and deleted files guide covers the wider version.
The short version
A drive that won't mount with a space manager invalid error almost always still has your files. Don't erase it, and don't keep poking it. Image it once, recover from the image, and keep the original untouched until the data is safely off.
General information only, not a guarantee. Recovery outcomes depend on the specific drive.
A drive that won't mount, and files that matter?
Tell us what happened and what is on the drive through the contact form. We'll tell you straight whether it's a clone-and-recover job or one to stop touching now.