unRAID Server release 4.5-beta2 available

[limetech]

Can you mount the flash drive with the noatime and nodiratime attributes enabled? 

 

Yes, meant to do that, thanks for the reminder 

[limetech]

I have no users defined. Security set to simple. For the life of me i cant connect to SAMBA as it prompts for a username and password.

 

I can not duplicate this.  If someone knows how to make it happen please post.  Also for the server that is doing this, please capture the output of this command:

 

testparm -sv

 

and include in your post.  Thanks!

[SSD]

I happened to look at my syslog today and noticed a large block of unexpected activity (about 2000 lines).  Here is the beginning.  Based on the timestamp, it seems this entire block was generated all at one time.

 

This occurred about 24 hours after booting, and about 12 hours after completing a full parity check (which ran successfully).

 

This log activity has not interfered with the system, and has not recurred.

 

Mar 9 19:05:08 Tower kernel: ad+0x38/0x5e
Mar 9 19:05:08 Tower kernel: [] kthread+0x0/0x5e
Mar 9 19:05:08 Tower kernel: [] kernel_thread_helper+0x7/0x10
Mar 9 19:05:08 Tower kernel: ksuspend_usbd S c04675bc 0 169 2
Mar 9 19:05:08 Tower kernel: c03ed1a0 00000046 00000001 c04675bc c046a780 c046a780 f744f600 f744f754
Mar 9 19:05:08 Tower kernel: c3f83780 00000001 c032618f f744d440 c04675bc c046a780 c046a780 f744f600
Mar 9 19:05:08 Tower kernel: f744f754 00000246 f74fd380 c012ad0d f74fd38c 00000000 c012ad94 00000000
Mar 9 19:05:08 Tower kernel: Call Trace:
Mar 9 19:05:08 Tower kernel: [] schedule+0x582/0x5d2
Mar 9 19:05:08 Tower kernel: [] worker_thread+0x0/0xbf
Mar 9 19:05:08 Tower kernel: [] worker_thread+0x87/0xbf
Mar 9 19:05:08 Tower kernel: [] autoremove_wake_function+0x0/0x2b
Mar 9 19:05:08 Tower kernel: [] worker_thread+0x0/0xbf
Mar 9 19:05:08 Tower kernel: [] kthread+0x38/0x5e
Mar 9 19:05:08 Tower kernel: [] kthread+0x0/0x5e
Mar 9 19:05:08 Tower kernel: [] kernel_thread_helper+0x7/0x10
Mar 9 19:05:08 Tower kernel: khubd S c04675bc 0 175 2
Mar 9 19:05:08 Tower kernel: c2895880 00000046 f74dbf14 c04675bc c046a780 c046a780 f744f960 f744fab4
Mar 9 19:05:08 Tower kernel: c3f7b780 00000000 00000000 c28d1400 c28d145c f77cb814 c02918c1 c28d1400
Mar 9 19:05:08 Tower kernel: 00000000 00000246 c0401b58 c28d1400 f74dbfbc f77cb814 c028b8f4 f74dbfa8
Mar 9 19:05:08 Tower kernel: Call Trace:
Mar 9 19:05:08 Tower kernel: [] usb_create_sysfs_dev_files+0x56/0x69
Mar 9 19:05:08 Tower kernel: [] hub_thread+0xa17/0xaa1
Mar 9 19:05:08 Tower kernel: [] schedule+0x582/0x5d2
Mar 9 19:05:08 Tower kernel: [] autoremove_wake_function+0x0/0x2b
Mar 9 19:05:08 Tower kernel: [] hub_thread+0x0/0xaa1
Mar 9 19:05:08 Tower kernel: [] kthread+0x38/0x5e
Mar 9 19:05:08 Tower kernel: [] kthread+0x0/0x5e
Mar 9 19:05:08 Tower kernel: [] kernel_thread_helper+0x7/0x10
Mar 9 19:05:08 Tower kernel: kseriod S c04675bc 0 178 2
Mar 9 19:05:08 Tower kernel: f777ddc0 00000046 c0405018 c04675bc c046a780 c046a780 f7518000 f7518154
Mar 9 19:05:08 Tower kernel: c3f7b780 00000000 c0404a28 c0405018 00000000 c027114e f74fd63c 00000000
Mar 9 19:05:08 Tower kernel: 00000000 00000246 00000000 f7504440 f751dfc8 00000246 c02a3f60 f751dfb4
Mar 9 19:05:08 Tower kernel: Call Trace:
Mar 9 19:05:08 Tower kernel: [] bus_for_each_dev+0x4d/0x56
Mar 9 19:05:08 Tower kernel: [] serio_thread+0x274/0x2d8
Mar 9 19:05:08 Tower kernel: [] autoremove_wake_function+0x0/0x2b
Mar 9 19:05:08 Tower kernel: [] serio_thread+0x0/0x2d8
Mar 9 19:05:08 Tower kernel: [] kthread+0x38/0x5e
Mar 9 19:05:08 Tower kernel: [] kthread+0x0/0x5e
Mar 9 19:05:08 Tower kernel: [] kernel_thread_helper+0x7/0x10
Mar 9 19:05:08 Tower kernel: kswapd0 S c04675bc 0 242 2
Mar 9 19:05:08 Tower kernel: f629e540 00000046 00000000 c04675bc c046a780 c046a780 f7518a20 f7518b74
Mar 9 19:05:08 Tower kernel: c3f83780 00000001 c0145e69 00000000 00000000 00000000 c04677bc 00000018
Mar 9 19:05:08 Tower kernel: c040ac80 00000246 00000001 c040ac80 c040b880 ffffffff c01466a0 c04677bc
Mar 9 19:05:08 Tower kernel: Call Trace:
Mar 9 19:05:08 Tower kernel: [] shrink_slab+0x12d/0x139
Mar 9 19:05:08 Tower kernel: [] kswapd+0xd3/0x497
Mar 9 19:05:08 Tower kernel: [] isolate_pages_global+0x0/0x1ef
Mar 9 19:05:08 Tower kernel: [] autoremove_wake_function+0x0/0x2b
Mar 9 19:05:08 Tower kernel: [] complete+0x28/0x36
Mar 9 19:05:08 Tower kernel: [] kswapd+0x0/0x497
Mar 9 19:05:08 Tower kernel: [] kthread+0x38/0x5e
Mar 9 19:05:08 Tower kernel: [] kthread+0x0/0x5e
Mar 9 19:05:08 Tower kernel: [] kernel_thread_helper+0x7/0x10
Mar 9 19:05:08 Tower kernel: aio/0 S c04675bc 0 283 2
Mar 9 19:05:08 Tower kernel: c03ed1a0 00000046 00000000 c04675bc c046a780 c046a780 f7518d80 f7518ed4
Mar 9 19:05:08 Tower kernel: c3f7b780 00000000 c032618f f744ca20 c04675bc c046a780 c046a780 f7518d80
Mar 9 19:05:08 Tower kernel: f7518ed4 00000246 f7733380 c012ad0d f773338c 00000000 c012ad94 00000000
Mar 9 19:05:08 Tower kernel: Call Trace:
Mar 9 19:05:08 Tower kernel: [] schedule+0x582/0x5d2
Mar 9 19:05:08 Tower kernel: [] worker_thread+0x0/0xbf
Mar 9 19:05:08 Tower kernel: [] worker_thread+0x87/0xbf
Mar 9 19:05:08 Tower kernel: [] autoremove_wake_function+0x0/0x2b
Mar 9 19:05:08 Tower kernel: [] worker_thread+0x0/0xbf
Mar 9 19:05:08 Tower kernel: [] kthread+0x38/0x5e
Mar 9 19:05:08 Tower kernel: [] kthread+0x0/0x5e
Mar 9 19:05:08 Tower kernel: [] kernel_thread_helper+0x7/0x10
Mar 9 19:05:08 Tower kernel: aio/1 S c04675bc 0 284 2
Mar 9 19:05:08 Tower kernel: c03ed1a0 00000046 00000001 c04675bc c046a780 c046a780 f75190e0 f7519234
Mar 9 19:05:08 Tower kernel: c3f83780 00000001 c032618f f778ca20 c04675bc c046a780 c046a780 f75190e0
Mar 9 19:05:08 Tower kernel: f7519234 00000246 f7733400 c012ad0d f773340c 00000000 c012ad94 00000000
Mar 9 19:05:08 Tower kernel: Call Trace:
Mar 9 19:05:08 Tower kernel: [] schedule+0x582/0x5d2
Mar 9 19:05:08 Tower kernel: [] worker_thread+0x0/0xbf
Mar 9 19:05:08 Tower kernel: [] worker_thread+0x87/0xbf
Mar 9 19:05:08 Tower kernel: [] autoremove_wake_function+0x0/0x2b
Mar 9 19:05:08 Tower kernel: [] worker_thread+0x0/0xbf
Mar 9 19:05:08 Tower kernel: [] kthread+0x38/0x5e
Mar 9 19:05:08 Tower kernel: [] kthread+0x0/0x5e
Mar 9 19:05:08 Tower kernel: [] kernel_thread_helper+0x7/0x10
Mar 9 19:05:08 Tower kernel: nfsiod S c04675bc 0 290 2
Mar 9 19:05:08 Tower kernel: c03ed1a0 00000046 00000000 c04675bc c046a780 c046a780 f7519440 f7519594
Mar 9 19:05:08 Tower kernel: c3f7b780 00000000 c032618f f744ca20 c04675bc c046a780 c046a780 f7519440
Mar 9 19:05:08 Tower kernel: f7519594 00000246 f7733d80 c012ad0d f7733d8c 00000000 c012ad94 00000000
Mar 9 19:05:08 Tower kernel: Call Trace:
Mar 9 19:05:08 Tower kernel: [] schedule+0x582/0x5d2
Mar 9 19:05:08 Tower kernel: [] worker_thread+0x0/0xbf
Mar 9 19:05:08 Tower kernel: [] worker_thread+0x87/0xbf
Mar 9 19:05:08 Tower kernel: [] autoremove_wake_function+0x0/0x2b
Mar 9 19:05:08 Tower kernel: [] worker_thread+0x0/0xbf
Mar 9 19:05:08 Tower kernel: [] kthread+0x38/0x5e
Mar 9 19:05:08 Tower kernel: [] kthread+0x0/0x5e
Mar 9 19:05:08 Tower kernel: [] kernel_thread_helper+0x7/0x10
Mar 9 19:05:08 Tower kernel: cifsoplockd S c04675bc 0 295 2
Mar 9 19:05:08 Tower kernel: c2a51c00 00000046 f744d440 c04675bc c046a780 c046a780 f75197a0 f75198f4
Mar 9 19:05:08 Tower kernel: c3f83780 00000001 c012562a f777ffa0 00859eed f777ffa0 00000286 c0125731
Mar 9 19:05:08 Tower kernel: 00000000 00000286 00859eed 00859eed 00000000 00000000 c0326315 f74668fc
Mar 9 19:05:08 Tower kernel: Call Trace:
Mar 9 19:05:08 Tower kernel: [] lock_timer_base+0x19/0x35
Mar 9 19:05:08 Tower kernel: [] __mod_timer+0x8d/0x97
Mar 9 19:05:08 Tower kernel: [] schedule_timeout+0x6b/0x86
Mar 9 19:05:08 Tower kernel: [] process_timeout+0x0/0x5
Mar 9 19:05:08 Tower kernel: [] cifs_oplock_thread+0x0/0x196
Mar 9 19:05:08 Tower kernel: [] cifs_oplock_thread+0x15f/0x196
Mar 9 19:05:08 Tower kernel: [] cifs_oplock_thread+0x0/0x196
Mar 9 19:05:08 Tower kernel: [] kthread+0x38/0x5e
Mar 9 19:05:08 Tower kernel: [] kthread+0x0/0x5e
Mar 9 19:05:08 Tower kernel: [] kernel_thread_helper+0x7/0x10
Mar 9 19:05:08 Tower kernel: cifsdnotifyd S c04675bc 0 296 2

[limetech]

I happened to look at my syslog today and noticed a large block of unexpected activity (about 2000 lines).  Here is the beginning.  Based on the timestamp, it seems this entire block was generated all at one time.

 

This occurred about 24 hours after booting, and about 12 hours after completing a full parity check (which ran successfully).

 

This log activity has not interfered with the system, and has not recurred.

 

 

Do you still have the complete log?  If so, please email to me: [email protected]

[SSD]

On its way.

 

Please advise whether this is something I should be concerned about.

 

There are no (appanet) negative side effects.  The array is still running fine.  I have not rebooted.

 

With the exception of this unexpected spurt of log messages, this version runs quite well.  I am not usually one to jump quickly on the latest beta, but did this time to verify to test out the queue_depth performance issue, and have seen no reason to downgrade.  The new kernal and Samba are working with no problem.  I do not use Active Directory, so can't comment on that part.  I do have user security implemented with my custom samba shares, and that is still all working fine.

[NAS]

I have no users defined. Security set to simple. For the life of me i cant connect to SAMBA as it prompts for a username and password.

 

I can not duplicate this.  If someone knows how to make it happen please post.  Also for the server that is doing this, please capture the output of this command:

 

testparm -sv

 

and include in your post.  Thanks!

 

based on the posts of another user i tried stopping the array. enabling user security . start array. stop array. enable simple security. start array.

 

From this point on I cannot replicate the bug anymore.

 

no clue what happened.

 

Edit: just rebooted and added a new disk. Still working as expected. weird

[erikatcuse]

So I'm running 4.5beta2 and the other day one of my drives was disabled!  I thought it was because I was testing my LSI card (it has been installed since 4.4 but not used and still isn't being used) but I don't think that is the problem.  So I did a SMART check and it passed.  I also checked the file structure to see if all my data was there and it was.  So I used the restore button.  The server came back on line and parity check started.  Everything was fine.  I moved the server to another location and upon reboot again the disk3 was disabled.  I went and check all the cable connections and it looked fine.  This time before I powered off I unassigned the disk and then rebooted and reassigned the drive to the same location and and it saw it as new so I did a Start to rebuild the drive from parity it started working fine.  However this morning the disk was disabled again.  I since have replace the disk and am rebuilding parity.

 

Now the disk that was causing the problems is in the system and I just mounted it using unmenu.  All my data is there and it seems to be working fine.  I've attached a sys log of last nights event.  I'm not sure what to make of it.

 

Any suggestions would be appreciated.

 

Erik

[RobJ]

I'm hoping that Tom will consider adding a hard disks forum in the Support area, just for hard drive problems, so people won't have to wonder where to put these, as they are generally unrelated to specific versions of unRAID or the kernel.

 

Early on when I started with unRAID, I had similar problems where a drive or controller would be dropped, and I would face the choice of either rebuilding parity or rebuilding the data drive.  Now, it is much simpler, use the Trust My Array procedure, which quickly restores the array to all green, and starts a parity check, which can be aborted after 2% to 5% (unless a user has not run it very recently).

 

In your case, the system is losing contact with Disk 3 (sde), and once the drive has been disabled, all kinds of errors are reported afterward, but can be ignored, as they are just because the drive was disabled by the kernel.  It first tried to slow down the access to it, to see if that would help, but lost complete access to it, and disabled the drive.  Even though the link came back up (sounds like a loose connection somewhere), it is too late, unrecoverable.  In cases like these, there is usually nothing at all wrong with the drive, just a serious problem with the interface to the drive.  The best information I can gather so far, for the error messages you had, is either a bad cable or bad or loose connector, or especially a faulty or not well seated backplane connection.  It could also be a problem with faulty power to this drive, so check the quality of the power cable to it, especially any power cable splitters involved.

 

I've begun working on a Wiki document to help users diagnose their own and others drive issues.  It is only a beginning yet, poorly formatted and organized, and very incomplete...

  The Analysis of Drive Issues

 

Something you may or may not be aware of, in your syslog, your LSI card is being recognized and set up, and a single channel (scsi1) was created for it, for drives on it.  I'm not familiar with ioc0, will have to read up on it, but I assume it is SAS drive related.  Here are the related lines:

Mar 15 01:03:53 Backup kernel: Fusion MPT base driver 3.04.07

Mar 15 01:03:53 Backup kernel: Copyright © 1999-2008 LSI Corporation

Mar 15 01:03:53 Backup kernel: Fusion MPT SAS Host driver 3.04.07

Mar 15 01:03:53 Backup kernel: mptsas 0000:01:00.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16

Mar 15 01:03:53 Backup kernel: mptbase: ioc0: Initiating bringup

  ...

Mar 15 01:03:53 Backup kernel: ioc0: LSISAS1068E B3: Capabilities={Initiator}

Mar 15 01:03:53 Backup kernel: mptsas 0000:01:00.0: irq 379 for MSI/MSI-X

Mar 15 01:03:53 Backup kernel: mptbase: ioc0: PCI-MSI enabled

Mar 15 01:03:53 Backup kernel: mptsas 0000:01:00.0: setting latency timer to 64

  ...

Mar 15 01:03:53 Backup kernel: scsi1 : ioc0: LSISAS1068E B3, FwRev=011a0000h, Ports=1, MaxQ=478, IRQ=379

[erikatcuse]

Thanks Rob

 

I'll check the connectors and since I'm rebuilding to a spare drive I have I'll see it does the same thing.

 

And yes it's SAS related...unraid doesn't recognize the drives because they start with a SAS instead of a SCSI identifier (a bug that I hoped was fixed in this beta but not.)

Using unmenu I can mount my SATA drive connected to the LSI card but that's it.  Tom's aware of this and hopefully it will be worked out soon.

 

 

[marcusone]

I can't get "simple" security to work either.  And since I don't have the pro version, I can't try the AD trick listed here.  Is there a fix yet?

 

Thanks,

[marcusone]

I can't get "simple" security to work either.  And since I don't have the pro version, I can't try the AD trick listed here.  Is there a fix yet?

 

Thanks,

 

Well, after some more playing around I got it to work (using some configs I've done on another samba server I have);

 

putting in /boot/config/smb.extra

workgroup = home
server string =
security = share
guest ok = yes
guest account = root
force user = root
force group = root

 

was the only way to get the correct workgroup setup (without the first line it would default to workgroup, despite what i put in the GUI)

and I had to add the lines to force user/group.

 

Not sure this is the best way.. but its currently working.

[marcusone]

Now that I got the simple shares working and had time to "play around" with this beta; I'm impressed with the speed improvement to say the least!

I get steady 15+MB/s on writes now!  A full 50+% increase (use to bounce around 7-15, with steady streams at 10).  It actually bounces between 15-20MB/s, so the average is likely higher than 15!

 

Great work!

[limetech]

Yeah bug in -beta2 where initial samba configuration is not completely initialized.  Workaround is to go to the Settings page and just click the 'Apply' button in the Identification section - no need to change change any settings (unless you want to).  Clicking 'Apply' in this case will update the boot/ident.cfg and boot/share.cfg files & restart samba.  Fixed in -beta3.

[marcusone]

Yeah bug in -beta2 where initial samba configuration is not completely initialized.  Workaround is to go to the Settings page and just click the 'Apply' button in the Identification section - no need to change change any settings (unless you want to).  Clicking 'Apply' in this case will update the boot/ident.cfg and boot/share.cfg files & restart samba.  Fixed in -beta3.

 

That didn't work for me... click apply, with or without changes, still required a login (even after rebooting my windows machine, which is sometimes needs to 'forget' what credentials you have tried to use on prior connections).  The only way I've been able to get it to work is by adding lines to the samba config file.

 

Hopefully you will be able to release beta3 soon

[SuperW2]

Not sure if this is a 4.5Beta 2 issue or not, but since the upgrade my Cache Mover is not function, either on the schedule or when I click "move now" (clicking does not initiate anything, my drives are spun down and I see no activity)... I have plenty of space on my Array.

[Nyago123]

Not sure if this is a 4.5Beta 2 issue or not, but since the upgrade my Cache Mover is not function, either on the schedule or when I click "move now" (clicking does not initiate anything, my drives are spun down and I see no activity)... I have plenty of space on my Array.

 

I have the same issue.  I don't know the specific cause.

 

 

[RobJ]

Not sure if this is a 4.5Beta 2 issue or not, but since the upgrade my Cache Mover is not function, either on the schedule or when I click "move now" (clicking does not initiate anything, my drives are spun down and I see no activity)... I have plenty of space on my Array.

 

I have the same issue.  I don't know the specific cause.

 

Your syslog (found here) named syslog 2009-03-09.txt (using unRAID v4.4.2) shows the mover lines set up with cron.  The syslog named syslog 2009-03-15.txt (using unRAID v4.5-beta2) does NOT have those lines.  Both show that User Shares is configured and setup, and the Cache drive is mounted and setup for caching.

 

You may want to email Tom.

 

SuperW2:  No ideas so far, as you DO have the cron mover lines, and I don't see anything wrong in your syslog.  I hate to ask the obvious question (don't shoot me!), but are you positive there was something there to be moved?

[Nyago123]

Will do... thanks Rob.  You went above and beyond pulling my syslogs from another thread. 

 

For the record, trying to run the mover manually does not work on 4.5-beta2 for me as well.

 

It also seems that over time, 4.5-beta2 gets more and more likely to hang temporarily (20-30 seconds) on a response to network requests (e.g. refreshing the main GUI page or browsing files from Windows Explorer).  This occurs either when the unRAID server is joined to an AD domain or not.  I've reverted to 4.4.2 and all is well again.

 

[SuperW2]

SuperW2:  No ideas so far, as you DO have the cron mover lines, and I don't see anything wrong in your syslog.  I hate to ask the obvious question (don't shoot me!), but are you positive there was something there to be moved?

 

Fair question RobJ, no shooting....but yep, I can browse to the Cache Disk and see actual new files that need to be moved (and hundreds of empty folders that I wish got removed when the mover happened)...

 

 

I looked at the cache disk and folders modified on 03/09 2AM (Desktop Backup Job) and after were not moved and my BZRoot and BZImage Update on my Flash was 03/08 at 2PM...likely that I didn't copy any files to my array between 2pm and 2am...at least that I could find any folders, empty or full for on the cache drive...so I'm pretty sure there is some correlation...

 

I could down rev back to the released ver to see if the mover functions again....but if it's a bug I'd like to help troubleshoot...

 

Wade

 

 

 

[JarDo]

...but yep, I can browse to the Cache Disk and see actual new files that need to be moved (and hundreds of empty folders that I wish got removed when the mover happened)...

 

I haven't tried any of the methods mentioned on this site, or the alternate software methods, but you might find this information interesting as it relates to removing empty directories.  Let us know if you do try these methods and how they work for you:

 

http://duramecho.com/ComputerPrograms/DeleteEmptyDirectories/index.html

 

It looks as if FSlint can be scripted if the perl or shell script methods won't work.

[Joe L.]

...but yep, I can browse to the Cache Disk and see actual new files that need to be moved (and hundreds of empty folders that I wish got removed when the mover happened)...

 

I haven't tried any of the methods mentioned on this site, or the alternate software methods, but you might find this information interesting as it relates to removing empty directories.  Let us know if you do try these methods and how they work for you:

 

http://duramecho.com/ComputerPrograms/DeleteEmptyDirectories/index.html

 

It looks as if FSlint can be scripted if the perl or shell script methods won't work.

 

This should work...

find /mnt/cache -depth -type d -empty -exec rmdir {} \;

[SuperW2]

SuperW2:  No ideas so far, as you DO have the cron mover lines, and I don't see anything wrong in your syslog.  I hate to ask the obvious question (don't shoot me!), but are you positive there was something there to be moved?

 

Fair question RobJ, no shooting....but yep, I can browse to the Cache Disk and see actual new files that need to be moved (and hundreds of empty folders that I wish got removed when the mover happened)...

 

 

I looked at the cache disk and folders modified on 03/09 2AM (Desktop Backup Job) and after were not moved and my BZRoot and BZImage Update on my Flash was 03/08 at 2PM...likely that I didn't copy any files to my array between 2pm and 2am...at least that I could find any folders, empty or full for on the cache drive...so I'm pretty sure there is some correlation...

 

I could down rev back to the released ver to see if the mover functions again....but if it's a bug I'd like to help troubleshoot...

 

Wade

 

 

 

 

Well, my Cache Drive was going to be full...so I down-reved back to 4.4.2 and the mover is now copying stuff over via a manual push-button "Move Now Operation"...

 

Wade

[marcusone]

...but yep, I can browse to the Cache Disk and see actual new files that need to be moved (and hundreds of empty folders that I wish got removed when the mover happened)...

 

I haven't tried any of the methods mentioned on this site, or the alternate software methods, but you might find this information interesting as it relates to removing empty directories.  Let us know if you do try these methods and how they work for you:

 

http://duramecho.com/ComputerPrograms/DeleteEmptyDirectories/index.html

 

It looks as if FSlint can be scripted if the perl or shell script methods won't work.

 

Just do this to delete empty directories

> find -depth -mindepth 2 -type d -empty -exec rm -f {} \;

 

adjust depth, or remove as needed (man find on a full linux box to learn more).

[mashles]

Mover not working for me either. didnt even notice until cache drive filled up. Back to 4.4.2 for me.

[Joe L.]

 

This should work...

find /mnt/cache -depth -type d -empty -exec rmdir {} \;

 

I get this when I run that command....

 

rmdir: /mnt/cache: Device or resource busy

That is OK.  it is a mount-point.  The top directory cannot be removed, even if empty, because the cache drive is mounted on it.

 

All your empty directories at that point, as it works from the "deepest level" directories first.