Disk not ejected properly

[Gregg2]

It's a Western Digital 2TB drive, Blue. I don't know it there is a model number. Nothing on it or in the pamphlet says "model number". It might be this: KB #3865

 

[rpmurray]

It's a Western Digital 2TB drive, Blue. I don't know it there is a model number. Nothing on it or in the pamphlet says "model number". It might be this: KB #3865

If it's a Western Digital drive then the model number should start with WD, e. g. WD5000AZLX, sometimes also called part number.

 

[Gregg2]

OK. The blue-on-blue made P/N look like PIN to me. So it's
WDBB7B0020BBL-XA50218
Otherwise known as 2TB Blue. 😉

 

[satcomer]

OK. The blue-on-blue made P/N look like PIN to me. So it's
WDBB7B0020BBL-XA50218
Otherwise known as 2TB Blue. 😉

It sound like a sleep function on the drive!

 

[driver49]

Well. How discouraging.

To find this thread is three years old and that the problem that we're discussing here has first surfaced years before that...?!?!

I recently replaced a Mac mini M1 with a Mac mini M4 and thought that that might solve the problem that creeped up too often on the Mac mini M1.

Alas.

The external drive is a SanDisk/GDrive 12-TB HDD that I have formatted into four partitions. And at least once an hour, if not more, I get the error message that all four drives have been ejected improperly.

I have replaced the cable a couple of times and that hasn't made any difference.

So it is comforting on the one hand to read that others have the same problem.

It is also aggravating to see that the problem has been around for a long time and Apple hasn't done a thing about it.

I see that this thread is more than three years old so I don't know if putting an entry here will bring it to the top for others to see, but if anybody's got any fresh ideas, other than those that are already presented in this thread I'm all ears.

Thanks,

--Paul S

 

[Ben J.]

The external drive is a SanDisk/GDrive 12-TB HDD that I have formatted into four partitions.

So, that's four HFS+ partitions - no APFS?

This has been a problem showing up in forums like this for as long as I can remember, and there's never been a final solution that I've ever seen; just people fixing it with replacing drives, cables, dasychaining etc.

I think you're in for a lot of troubleshooting, if you're to find what's the issue in your specific situation - like, start with one HFS+ partition only on the HDD, no hubs, nothing else connected, maybe run the mac in safe mode for a while.

If it happens on both the M1 and M4 with the same drive, this points to the drive being the issue. Try a different HDD if you can.

 

[driver49]

Not that I fully understand the difference between HFS+ and APFS – though I know that Apple is now using the latter for all of its operating systems – were I to make sure the drive in question is formatted APFS would that help for starters? It is possible that it is not, as my use of this drive I believe predates Apple's shift to APFS. I will have to check. That would give me the option of putting all the data from the four partitions into one, though I would prefer not to have do that as they all have rather historical purposes for me. Do you think I would be able to use for partitions if everything is APFS? It also occurred to me that in addition to the Time Machine backups that aren't working. I also have Backblaze for off site back up (which is why I'm not panicking over this), I guess I would have to do those back ups over too. The priority here is to get the drives and the data to just stay connected to the damn computer. Thanks, --PS

 

[Ben J.]

Apple created APFS to accommodate the way SSDs work, better than MacOS Extended (HFS+), basically - but HFS+ is still supported. But from Tahoe and onwards, macOS must be run on a APFS formatted drive.

Both APFS and HFS+ can be used on both SSDs and HDDs, but my experience is, and most reports are, that SSDs work better with APFS, and HDDs are less stressed with HFS+.

APFS came with some very nice new features compared to HFS+;
Volumes:
Kinda like partitions, but can be created in seconds, and don't have to have a fixed size as partitions do - thus, they 'share the free space'.
Snapshots:
You can create a snapshot on a APFS volume at any time, and it's like a copy of the files directory of a volume - not the files themselves, so they're small in size. Any file/folder present in any snapshot will never be overwritten - even if you drag it to the trash and empty it and they seem to be gone from Finder. That way, any file, in any snapshot will stay (invisible) on the drive, and be restorable. Once you delete your snapshots, the space used by those 'invisible' files will become available again.

I don't, in any way, have a deep understanding of these these things, but what I've written above, I'm pretty sure of, and is some of what I've learned over the last decade of going from Intel macs and HFS+ to M-series ARM and APFS.

(Btw, Any drive can still be partitioned, and partitions can be a mix of HFS+ and APFS.)

 

[Gregg2]

Well, this problem just went away for me. I haven't seen that message since I last posted to this thread! I hope there's some good luck left in it! 😉

 

[driver49]

Well, this problem just went away for me. I haven't seen that message since I last posted to this thread! I hope there's some good luck left in it! 😉

Well, Lucky you! I have little faith/hope I'll have a similar experience.

 

[driver49]

I have been pondering what approach to take to finding a solution to this issue.

I am pretty sure the drive(s) in question are NOT AFPS.

Assuming that the big drive with the 4 partitions needs to be APFS, I have thought that I would get a new drive, format that accordingly, and then move all the data (~6TB) to the new one and see if THAT keeps getting kicked out.

But then it dawns on me: If I try to connect the "old" drive to the "new" drive through the Mini... I still have the problem of the drive(s) being kicked out – maybe dozens of times – during that copy/transfer.

I dunno, I guess I could try connecting both drives to my laptop and copy/transfer everything that way but that presents its own issues, like keeping a computer I wander around with tied to my desk for as long as it takes.

I know... first world problems.

Thanks for all the input here, it's very helpful.

--Paul S. (alias "driver49")

 

[Brian33]

@driver49 , I'm sorry to hear you've had this issue. I had it on Monterey with two Thunderbolt 2 connected HDD drives, occasionally overnight while the computer was sleeping. I put a lot of time into reading others' reports, and a lot of effort into troubleshooting, but never found a solution other than preventing my Mac from sleeping.

Luckily (and I do think luck is involved, as I believe it's some sort of race condition bug in macOS), I haven't had the issue since I got a new Mac running Sequoia (same drives, same cables). <shrug>

But then it dawns on me: If I try to connect the "old" drive to the "new" drive through the Mini... I still have the problem of the drive(s) being kicked out – maybe dozens of times – during that copy/transfer.

You could work around disconnects by NOT using Finder's drag-and-drop. Try either 'rsync' (command-line) or Carbon Copy Cloner (I've heard there's a free trial period). With either of these tools, you can re-start the copy and it will pick up from where it was interrupted. So, you might have to manually restart the process, but since it won't re-copy what's done, you'll eventually finish the job.

With 'rsync' it is a bit tricky to decide which options to use. Assuming you're running Sequoia or later, I recommend trying rsync -aHES <source> <dest> , where the -E will ensure copying extended attributes and ACLs (if any). See man rsync for the gory details. Oh yeah, use the --dry-run option at first to test the command.

Carbon Copy Cloner (or probably, Super Duper) would be a GUI way to do the job; easier for most.

 

[driver49]

Given my experience of late with Time Machines backups (discussed here), I am already thinking of using carbon copy cloner, which looks like a very robust program that offers a lot of features that Time Machine does not. So that's a useful suggestion except for the part where if the drives do kick out, I might have to sit there for the entire time it takes to transfer 6 TB of data between two drives. So much for getting it started and leaving it overnight. It's always something.

Thanks,

--Paul S.

 

[Ben J.]

You could work around disconnects by NOT using Finder's drag-and-drop. Try either 'rsync' (command-line) or Carbon Copy Cloner (I've heard there's a free trial period).

I strongly support the CCC recommendation. 30-day free trial but very much worth the money. Been using it for over a decade. Very solid, and great support from Mike Bombich and his team.

CCC is a very basic cloning software; it compares one drive (or volume, or folder) to another and simply uses basic macOS commands to copy files (or delete them) to make sure the two are identical. It also offers lots of extra features, of course, like creating and managing APFS snapshots, so the cloning abilitlty and the snapshots features together makes it a good solution for backups. (For HFS+ drives that lack snapshot support, it has Safetynet, that functions somewhat similarly.)

As Brian33 mentions, it can be used for other things than strictly backup; for example I have a couple of CCC 'tasks' that clones a folder on my drive to a folder on my googledrive at set times.

 

[hobowankenobi]

While less pervasive, we used to see some USB drives simply not stable a few OSes back. It appeared to be the USB chipset was just not compatible with the most recent MacOS...other than trying a different drive, never found a true fix. A driver could likely have fixed it...but that would essentially be Apple fixing/patching their OS, or a hardware driver from the chipset OEM, which are exceedingly rare.

Other backup options that have been good for years/decades (if anybody wants CCC alternatives):

Get Backup Pro
SuperDuper!
ChronoSync
SmartBackup (simple, but free)

 

[driver49]

I have been posting on a related thread here, FWIW, and just reported that my simultaneous issues appear to have crossed the threshold to solution.

A tech-support call with AppleCare yesterday after my backup problems were taken to engineering led us to running first aid on the source drive in question where we found some issues with two of the partitions.

Fortunately, I had a brand spanking new hard drive at my disposal so I copied all the data from the old source drive to a new one. Fortunately, there were no ejections while I transferred that 6 TB of data overnight.

I'm still a little concerned that I did see one spontaneous ejection this morning but other than that things seem stable and I'm getting a fresh backup done from the new source drive.

Thanks again for the input and support.

--PS

 

[monokakata]

I get random ejections on a 5 tb USB3 drive -- one that replaced one that had that problem. Usually one or two a day. It's a CCC backup drive, in drive-imaging mode. My other CCC drive, with snapshots, has never ejected offly. Different enclosure, different cable...the only common thing is the USB port on my Studio. So I'm pretty sure the port/chip is the problem and I'm done trying to do anything about it.

 

[MayflyMaven]

The CCC website had a blog post about this exact subject just recently:

"Disk Not Ejected Properly" 😡: what that macOS alert really means​

I have a USB hub that reliably reproduces this behavior. It's not just hubs, though, I've seen it all over the board. I can't remember the brand, but I had some early nVME storage device that used to exhibit this behavior any time it really got cookin'. I remember working around the annoyance by placing it on a cold pack. I had a horrid bout of this behavior maybe 8 years ago on a headless mini that had three USB enclosures attached to it (same make/model of enclosure, a mix of Seagate and Western Digital HDDs inside). Two of the enclosures worked just fine, one of them constantly dropped offline. I tried swapping disks between enclosures, swapping cables and ports; the issue always followed the enclosure.

 

[PaulD-UK]

Quote from that link:

"How do I stop the "Disk Not Ejected Properly" message on my Mac?​

There's no macOS setting that silences it, because it's a symptom rather than a preference -- the fix is to address whatever is making the drive disappear.
Start with the cheap, easy suspects: try a different cable, a different port..."

Specifically: NOT another port on the Mac, as that will just replicate the problem.
Instead USE A USB-C CONTROLLER PORT that is USING the same sort of Controller hardware as the (disconnecting) external hard drive/SSD is using - i.e. NOT designed by Apple.

The easiest way to connect a RealTek (or whoever) USB 3.* controller to your Apple Silicon Mac is to use a Thunderbolt 3 dock.
Do NOT use TB4 or TB5 docks, as they pass the external devices USB 3.* signals back to the host Mac controller chip (so the problem remains).

Only a TB3 dock uses it's own USB 3.* controller, and passes the USB 3.*data to and from the Mac using the PCIe protocol.
Which the Mac deals with impeccably, because PCIe TB3 on a Mac 'just works'...

So connect any USB 3.* devices. - HDs or SSDs - to a 10Gbps port on the TB3 dock, which will have its own PSU to keep any external devices reliably powered as required.

Apple understands reliable communication using PCIe to an external Intel TB3 controller chip...
Intel understands reliable communication to a PC-manufacturer USB 3.* controller...
RealTek (or whoever) understands reliable communication with USB enclosures and HD/SSD manufacturer's devices...

So unexpected drop-off error messages will most likely be a thing of the past (my experience).
Unless the firmware of your external enclosure, or HD (SSD) has been hardwired by the manufacturer to go to sleep at all possible occasions, in which case, they may just be incompatible... 🙁

Nitty gritty details:
If you want to connect lots of USB 3.* SSDs, and maintain the highest bandwidth to them simultaneously, then probably the best TB3 dock is Caldigit's TS3 dock.

That uses the older Intel Alpine Ridge TB3 controller chip, which allows more of TB3's PCIe data bandwidth to be routed to USB 3.* ports.

In about 2018 Intel changed over to the Titan Ridge TB3 chipset, which allows dual external 4K monitors to be connected.
This restricts the PCIe bandwidth available to USB 3.* ports to 10Gbps max between all the USB 3.*ports.
However, being a later chipset, it is likely more reliable than the older Alpine Ridge docks.

Good condition second user docks will be fine for this purpose.
My recommendation is to get one of the last WD D50 Game docks, that are still being sold new at reduced prices in some outlets (search around).
You can get this as a plain dock, or with a 1TB or 2TB SSD internally.

 

[hobowankenobi]

...I tried swapping disks between enclosures, swapping cables and ports; the issue always followed the enclosure.

That was my experience too. USB board, or perhaps port seemed to be the root of the issue. Not the drive itself.