dlandon

Do the following: Stop the array. Unassign the parity drive. Click the double arrows on the preclear UI. This will remove the resume file so a preclear will not restart. Re-assign the disk back to the array. You should be able to delete the historical entry. Start array. The preclear is resuming itself on array start and it should not if it's assigned to the array. I'll have to look into how this can be prevented.

There have been some issues with nvme disks not seating properly. If this is an nvme disk, power down, remove the device and re-install. See if that fixes the issue.

UD should be able to clear the disk.

The only thing I see is you are putting things in the recycle bin by machine (%m). That should have no bearing on whether or not the recycle bin works, but I'll do some testing. Are these diagnostics from when it is not working? Everything looks good. The next time you reboot and the recycle bin doesn't seem to work, post the diagnostics before doing anything else.

I created an exclusive share and deleted a file and it worked as expected. Post your diagnostics so I can have a look at your particular situation.

Sounds like a corruption in your flash drive. Glad you sorted it out.

Are you talking about a root share recycle bin?

The recycle bin only works at the share level. You can't create separate recycle bins for each user. That being said, you can have the one recycle bin work by user or machine. Click on the "Recycle Bin parameters" and read the help. You can add a parameter to have the recycle bin work by user or machine.

I ran your script on one of my test disks and the script ran the unmount command even with errors. Because it is related to one specific disk of yours, and I have no way of reproducing the issue, there isn't much I can do. Does the disk do the same thing on a different server?

No. If you are looking for a way to find UD devices, there are several ways. What is it you are looking for?

I think you might be asking in the wrong place. Try the support for the auto fan plugin.

The exclude really doesn't work because UD devices come and go and there is no way to 'exclude' them. I'm working on having cache_dirs changed so it only allows include shares, and not exclude shares. Glad it got sorted out for you.

You have 'stat' script errors - files not found. Fix those and the script will probably work. If you can't fix those comment them out and see if the unmount works.

Yes, it caches all mounts unless you set it up not to. You should only have cache_dirs set to include array and pool shares you want to be cached so it will ignore everything else.

I thought I found something, but I was working with an encrypted disk that is differnet. So, back to your issue. I don't understand how you can have different results when the disk is hot plugged and the lightning icon is used. The script is executed in exactly the same way.

Hold, I may have found something.

The remount is only a work around. We will have to come up with a solution.

'df' shows it's mounted: Filesystem Size Used Avail Use% Mounted on rootfs 2.9G 299M 2.6G 11% / tmpfs 32M 4.0M 29M 13% /run /dev/sda1 15G 993M 14G 7% /boot overlay 2.9G 299M 2.6G 11% /lib overlay 2.9G 299M 2.6G 11% /usr devtmpfs 8.0M 0 8.0M 0% /dev tmpfs 2.9G 0 2.9G 0% /dev/shm tmpfs 128M 304K 128M 1% /var/log tmpfs 1.0M 0 1.0M 0% /mnt/disks tmpfs 1.0M 0 1.0M 0% /mnt/remotes tmpfs 1.0M 0 1.0M 0% /mnt/addons tmpfs 1.0M 0 1.0M 0% /mnt/rootshare /dev/md1p1 11T 4.0T 7.0T 37% /mnt/disk1 /dev/sdb1 3.7T 992G 2.7T 27% /mnt/download_drive /dev/nvme0n1p1 477G 21G 456G 5% /mnt/unraid_files shfs 11T 4.0T 7.0T 37% /mnt/user0 shfs 11T 4.0T 7.0T 37% /mnt/user /dev/loop2 20G 5.6G 14G 29% /var/lib/docker /dev/sdd1 932G 94M 932G 1% /mnt/disks/transport data The log shows it was mounted, but no unmount was ever initiated: Feb 18 11:14:01 mitchsserver unassigned.devices: Successfully mounted '/dev/sdd1' on '/mnt/disks/transport data'. Feb 18 11:14:01 mitchsserver unassigned.devices: Device '/dev/sdd1' is not set to be shared. Feb 18 11:14:01 mitchsserver unassigned.devices: Running device script: 'transport_data.sh' with action 'ADD'. Feb 18 11:14:21 mitchsserver emhttpd: WDC_WDS100T2B0A-00SM50_200802A003DE (sdd) 512 1953525168 Feb 18 11:14:21 mitchsserver emhttpd: read SMART /dev/sdd These are the last lines in the log.

Not in the diagnostics you posted: Filesystem Size Used Avail Use% Mounted on rootfs 2.9G 314M 2.6G 11% / tmpfs 32M 4.1M 28M 13% /run /dev/sda1 15G 993M 14G 7% /boot overlay 2.9G 314M 2.6G 11% /lib overlay 2.9G 314M 2.6G 11% /usr devtmpfs 8.0M 0 8.0M 0% /dev tmpfs 2.9G 0 2.9G 0% /dev/shm tmpfs 128M 468K 128M 1% /var/log tmpfs 1.0M 0 1.0M 0% /mnt/disks tmpfs 1.0M 0 1.0M 0% /mnt/remotes tmpfs 1.0M 0 1.0M 0% /mnt/addons tmpfs 1.0M 0 1.0M 0% /mnt/rootshare /dev/md1p1 11T 4.0T 7.0T 37% /mnt/disk1 /dev/sdb1 3.7T 991G 2.7T 27% /mnt/download_drive /dev/nvme0n1p1 477G 21G 456G 5% /mnt/unraid_files shfs 11T 4.0T 7.0T 37% /mnt/user0 shfs 11T 4.0T 7.0T 37% /mnt/user /dev/loop2 20G 5.6G 14G 29% /var/lib/docker

The notification in the upper right is under the control of your script and is only a notification. If the mount button shows mounted after your script has completed, that's different. It's this code that shows the device is unmounted: /usr/local/emhttp/webGui/scripts/notify -e "Unraid Server Notice" -s "Unassigned Devices" -d "Device unmounted" -i "normal" This displays in the notifications in the upper right of the UI after the disk is unmounted. ls -la /mnt/disks/ Look for the mount point. If it's missing, the device is unmounted.

Your diagnostics shows the disk is unmounted. Does the UI show it is mounted?

The diagnostics you posted earlier did not show any of that activity. Can you post diagnostics after the unmount issue you are talking about?

Change the 'umount $DEVICE' to 'umount $UD_DEVICE' and see if it makes a difference. The $UD_DEVICE is the 'devX' designation of the disk.

That suggests a problem with the USB port, cable, or disk.

Post diagnostics after all that. Also, try detaching the disk, then re-attaching the disk. This simulates removing the disk and re-installing it. There are different sections of code in the script that get executed.