The disk wasnt ejected because one or more programs may be using it.

[zastopur]

Hi everyone!


MacBook Air m1 8/256
Sequoia 15.1
External nvme ssd Kingston NV2 [SNV2S/1000G], APFS
Ugreen 40gbps ssd enclosure
Connecting directly to the laptop by included with the enclosure cable


This issue was present before on Sonoma or Ventura, but appeared like 2-3 times within 6 months. It didn’t bother me at all. After I updated to Sequoia I basically haven’t been able to eject my drive without this error.

Im trying to eject from finder sidebar, from locations, from desktop - nothing.

I force quit all applications - nothing.


Spotlight index
At first an error with spotlight index caused this, but it fixed itself somewhere in the process of rebuilding the index after i reinstalled Sequoia from recovery. [I left my macbook to do all its post-update things for 10 hours, it rebuild the index and synchronized everything].


QuickLook
The same error, now caused by quicklook processes. Found a solution:
1. «lsof | grep /Volumes/drive1».

• If it says something about quicklook, I do «killall -9 QuickLookSatellite» or «/usr/bin/killall -KILL QuickLookUIService»

Finder
At this point ive already spent hours dealing with this problem. Found the solution - «killall -KILL Finder». It helped.


Restating
Restarting the Macboook without ejecting the drive. When I log in and quickly open finder to eject from sidebar I get the error, but then the drive successfully ejects.


Today
Today was different because nothing from above helped.

Ive also tried:
1. «diskutil unmount /Volumes/drive1»

• terminal says «Volume drive1 on disk5s1 failed to unmount»

2. Disk utility mount / unmount button is non-intractable with «Physical disk», intractable with «Container» and «Volume». Im pressing unmount on both: Failed to unmount “drive1” because it is currently in use.


Some more info

• Disk Utility First Aid didnt find any issues
• There are no other drives connected
• Drivedx doesnt show any issues with the drive - everything is 100% except from temperature, but temperature is still ok. There are no error logs.
• when i reconnect the drive after ejecting it, i wait and try to eject it again. sometimes the drive ejects successfully the second time, sometimes it doesn’t. most of the time this error appears when the drive has been connected for 10-30-60min or more.
• before this issue appeared my drive was disconnecting after macbook went to sleep. i guess the laptop just stops giving it power. but now even after leaving it sleep for a couple of hours the drive didnt disconnect. no idea if this has anything to do with the issue.
• When unable to eject system logs say:

Kernel - disk5s1: device is not readable.
Diskarbitrationd - unmounted disk, id = /dev/disk5s1, ongoing.
Diskarbitrationd - unmounted disk, id = /dev/disk5s1, failure.
Diskarbitrationd - unable to unmount /dev/disk5s1 (status code 0x00000010).
Kernel - print_unknown_ioctl:17066: disk5s1 unknown ioctl _IOW('X', 2, 48) (0x80305802), pid = 4143


Dont tell me to Force Eject.
The ONLY thing that ALWAYS helps is to turn off and then physically disconnect the cable. But doing this tens of times per day is driving me crazy.


ChatGPT
It helped before, so here is what it recommended that i havent tried yet. [I have no idea what is single-user mode, sip or disk identifier. If you think i shouldnt do something from this list please tell me].

1. Reboot in single-user mode, then “/sbin/mount -uw /diskutil unmountDisk /dev/disk5”
2. Try different cable / different port
3. Check and disable SIP: Boot into recovery, then “csrutil status”, then disable it.
4. Unmount using Disk Identifier: “sudo diskutil unmountDisk force /dev/disk5”
5. Reset NVRAM
6. Connect the drive to another macbook to see if the issue is drive-specific

Im not experienced. Everything that I’ve tried has been found on the internet, this forum specifically (thanks to everyone here btw!), and also with ChatGPT.

The main frustration is: each time i see this error its for a different reason. And i have a big list of troubleshooting steps to go through EACH TIME to try and fix it. “Just works” they said …

If you have any questions, want any additional info or logs, screenshots - please ask, i will provide anything you need.

edit: also why doesnt macOS just kill its own processes that don't allow the user to do what he wants to? is that too much to ask for? why do I need to open terminal to know what is using my drive, cant you just say like at least the name of the process?

edit2: I THINK I FOUND IT
It appears that the issue is still with spotlight index. "mds_store" is using files on the drive for indexing. but I don't see the index progress bar in spotlight. it appears for a couple of seconds with a light-blue color and disappears. I've already rebuilt the index 2 times. Maybe spotlight index is stuck on some files, but how do I know on which ones? I've read somewhere that spotlight finally finishes index after a couple of weeks. so, should I wait and see? should I rebuild the index again? how do I know which files the index is having a problem with?

btw thank you to everyone who replies ! appreciate it!

 

[Ben J.]

What is "drive1"?
A whole drive? A APFS volume that you created?

I had a problem with a tiny 10MB partition on an external SSD that would never eject. If you tried, it would immediately pop up again. The only thing that would work was to right-click the volume in Disk Utilities and select 'deactivate'. I suspect this was some kind of "unerasable" partitions that drive manufacturers sometimes put on their drives. I made an applescript to run that command at startup, and I don't have to see it again.

Unsure if this is because of some security settings I changed, but my Finder is 'quittable' FWIW.

It was a long time since I worried about the "Disk was improperly ejected" warning.

 

[Bigwaff]

Happen when MBA using battery only or connected to AC power, or not matter? Happen when writing data to disk or when disk appear idle, or not matter?

 

[zastopur]

What is "drive1"?
A whole drive? A APFS volume that you created?

I had a problem with a tiny 10MB partition on an external SSD that would never eject. If you tried, it would immediately pop up again. The only thing that would work was to right-click the volume in Disk Utilities and select 'deactivate'. I suspect this was some kind of "unerasable" partitions that drive manufacturers sometimes put on their drives. I made an applescript to run that command at startup, and I don't have to see it again.

Unsure if this is because of some security settings I changed, but my Finder is 'quittable' FWIW.

It was a long time since I worried about the "Disk was improperly ejected" warning.

"drive1" is an apfs volume. What is "deactivate" in Disk Utility. First time hearing about some unreadable manufacturer partitions. When I bought the drive a formatted it with macOS. Thank you for your reply!

 

[Grumpus]

Could it be that Spotlight indexing is keeping the external drive busy? Try adding the drive to the Spotlight Privacy panel in System Settings.

 

[zastopur]

Happen when MBA using battery only or connected to AC power, or not matter? Happen when writing data to disk or when disk appear idle, or not matter?

Doesn't matter - on battery and when charging / connected to power. Sometimes the problem is because of spotlight or quicklook processes using the drive. But yesterday nothing was using the drive, but I was still unable to eject it. Thank you for your reply!

 

[zastopur]

Could it be that Spotlight indexing is keeping the external drive busy? Try adding the drive to the Spotlight Privacy panel in System Settings.

Ive already rebuild the index, multiple times. Spotlight doesnt cause this issue anymore after I rebuilt the index. It is either because of quicklook, or for some other unknown reason. Again - nothing is using the drive. No indexing, no programs / processes using the drive (at least that's what "lsof | grep /Volumes/drive1" says), no files / programs opened - nothing. Thank you for your reply!

 

[eSaF-Eddie]

One of the culprits for this behavior is Time Machine or a Backup Drive.

 

[Grumpus]

Ive already rebuild the index, multiple times. Spotlight doesnt cause this issue anymore after I rebuilt the index. It is either because of quicklook, or for some other unknown reason. Again - nothing is using the drive. No indexing, no programs / processes using the drive (at least that's what "lsof | grep /Volumes/drive1" says), no files / programs opened - nothing. Thank you for your reply!

You need to be using sudo lsof to see everything.

 

[zastopur]

Thanks! Will try

You need to be using sudo lsof to see everything.

 

[zastopur]

it is spotlight again, "mds_store". gonna rebuild the index again

 

[Grumpus]

it is spotlight again, "mds_store". gonna rebuild the index again

Consider whether you really for Spotlight to index the drive, especially if it's not connected all the time. Adding the drive to the Spotlight Privacy panel in System Settings will delete the current index from the drive and prevent the drive from being indexed in the future (or it's removed from the Privacy panel). You should be able to drag the drive icon to the Privacy panel to add it.

 

[Ben J.]

Consider whether you really for Spotlight to index the drive, especially if it's not connected all the time. Adding the drive to the Spotlight Privacy panel in System Settings will delete the current index from the drive and prevent the drive from being indexed in the future (or it's removed from the Privacy panel). You should be able to drag the drive icon to the Privacy panel to add it.

…and if a volume is removed and re-added to that panel, it will be reindexed by Spotlight.

 

[Boneslasher]

I'm glad I found this post as I have been having this issue ever since upgrading to Sequoia (15.1) from Sonoma (14.7).
The internal SSD is using mds_stores all day but at around 100-150% CPU, all Efficiency Cores and I am using AppTamer to keep it in its ranks when I need to save battery. Shutting down, then booting in Safe Mode, waiting a couple of minutes, then restarting normally gets rid of the issue for... a few hours, then it comes back.
The problematic SSD is an external Crucial X8 2TB with two volumes, one with a 1.25 TB quota dedicated to my Time Machine and the other with my GOG Library and a few other minor things. I connect the disk about once a day to perform the TM backup, which automatically mounts both volumes. TM backup performs flawlessly and then I hold Option to eject both volumes at the same time. The TM volume is ejected immediately while the other one doesn't want to collaborate. Sadly, I have had to Force Eject it multiple times in the last week.
I honestly have no idea what to do: today I tried to keep it connected for 12 hours straight and it just wouldn't complete the process. Even if I could add this volume to the spotlight exclusion, I do not think this should be the solution, since Sonoma never gave any problem.
Could it be the new QuickLook thing? May macOS be trying to append the quicklook preview to each file of the GOG Library? 100s of 1000s of smallest files? If so, is there a way one could prevent it?

 

[zastopur]

…and if a volume is removed and re-added to that panel, it will be reindexed by Spotlight.

I know and I've rebuilt the index that way one time already and also rebuilt it via terminal. I suppose after reinstalling sequoia the system did index rebuild by itself. gonna try that one more time if the issue doesnt go away in a cup of weeks

 

[zastopur]

I'm glad I found this post as I have been having this issue ever since upgrading to Sequoia (15.1) from Sonoma (14.7).
The internal SSD is using mds_stores all day but at around 100-150% CPU, all Efficiency Cores and I am using AppTamer to keep it in its ranks when I need to save battery. Shutting down, then booting in Safe Mode, waiting a couple of minutes, then restarting normally gets rid of the issue for... a few hours, then it comes back.
The problematic SSD is an external Crucial X8 2TB with two volumes, one with a 1.25 TB quota dedicated to my Time Machine and the other with my GOG Library and a few other minor things. I connect the disk about once a day to perform the TM backup, which automatically mounts both volumes. TM backup performs flawlessly and then I hold Option to eject both volumes at the same time. The TM volume is ejected immediately while the other one doesn't want to collaborate. Sadly, I have had to Force Eject it multiple times in the last week.
I honestly have no idea what to do: today I tried to keep it connected for 12 hours straight and it just wouldn't complete the process. Even if I could add this volume to the spotlight exclusion, I do not think this should be the solution, since Sonoma never gave any problem.
Could it be the new QuickLook thing? May macOS be trying to append the quicklook preview to each file of the GOG Library? 100s of 1000s of smallest files? If so, is there a way one could prevent it?

after installing a macOS update it rebuilds the index, first it rebuilds the system disk then external ones. right now I think ill wait for 1-2 weeks to see if the issue disappears. I've read on forums that sometimes it can take days and weeks for spotlight to fully rebuild the index. if the issue is still present after those weeks ill rebuild the index the apple recommended way (by adding the drive to system settings spotlight privacy and removing the drive from there, this procedure rebuilds the index).

during those weeks ill look into spotlight index more. maybe there is a way to know if it is having problems with certain files / directories, maybe there are some broken files, unreadable files, files that need some clearance to access or something.

 

[Boneslasher]

What helped me in the end was going to Console, filter processes by MDS, see how there was at least one error per second (so 60+ / minute). I opened Activity Monitor, found MDS_WORKER process and Force Quit it.
After a short while, macOS relaunched it and, at this point, most errors were gone, Spotlight resumed operating beautifully, and I could also eject the drive naturally (without losing its index, by the way!).
I hope this may help you as well.

 

[Daverich4]

In my case it certainly has something to do with Spotlight. I haven't been able to eject any of my external drives since upgrading to Sequoia. Just for the heck of it I added one of the drives to Spotlight Privacy and now it ejects without any error messages. Not sure what to do next to get Spotlight to work properly but for now I'm just going to disable it on my external drives.

 

[ethandt72]

In my case it certainly has something to do with Spotlight. I haven't been able to eject any of my external drives since upgrading to Sequoia. Just for the heck of it I added one of the drives to Spotlight Privacy and now it ejects without any error messages. Not sure what to do next to get Spotlight to work properly but for now I'm just going to disable it on my external drives.

THANK YOU FOR THIS TIP!

I just spent an hour and a half on the phone with Western Digital tech support about this issue, which is happening with multiple drives, one HDD and one SSD. We ended up figuring that it was the update to 15.1 that did it, but no clue how to fix it. This worked, so thank you very much, just tried it with both of my drives successfully.

 

[nicolaselhani]

Glad I found this thread. Just upgraded from a 2013 MBP with no issues to an M4 Max with a couple of external SSDs all giving me the "Force Eject" issue. I thought the issue were the drives I had, cable, maybe their controller hardware being defective or something.

After finding this thread, I tried adding them to Spotlight Privacy with no success, had to restart a couple of times before I was able to. My question is in the future if I add new drives and forget to directly add them to Spotlight Privacy, is it generally considered safe to kill the mds_stores process in Activity Monitor to then be able to Eject the drive instead of having to do a whole restart of the mac? Would that negatively impact local process or any data on the ssd related to building an index? Thanks. what a crazy bug

 

[auxbuss]

I was having this problem. In fact, it caused two kernel panics. Ejecting with `diskutil` solved the panics. However, in my case, the latest Sequoia release, 15.1.1, has fixed it.

 

[zastopur]

a little update: the issue doesnt appear as often as it was. im running the latest 15.1.1 right now, maybe updating fixed it (auxbuss seems to have the same experience). but I still see the error like a couple of times per week. ive created an iOS shortcut that kills all mds processes and use it when unable to eject. I agree with nicolasenhali - im not at all a fan of force quitting system processes, so im gonna try Boneslasher's method of force quitting only mds_worker when unable to eject. once again big thanks for everyone who replies!

 

[!!!]

Unmount using Disk Identifier: “sudo diskutil unmountDisk force /dev/disk5”

You don't typically need sudo, and you can just drag and drop the volume into terminal instead of finding the /dev/disk number.

It's a better solution than force-quitting mds (which should be fine, but I've seen that corrupt the index and cause a complete reindex to happen)

 

[SaguaroSeven]

Very irritating... this problem still exists on 15.1.1. Plugged in a new USB SSD which i formatted APFS encrypted. Couldn't eject it as described here; mds was the culprit. When I checked Settings > Spotlight > Privacy the newly formatted disk was shown in the list, so WFT was Spotlight doing? Wierd.

 

[T Coma]

Same issue here with a Crucial X6 used for external Photos library on an M1 MBP. It's on the Spotlight Privacy list, and still cannot ever eject, but just this disk. I either have to force eject or remove after shutdown. Sequoia 15.2