JorgeB

That's not the problem, looks like an UD issue, you should post on the plugin support thread: Oct 4 16:01:08 Tower unassigned.devices: Successfully mounted '/dev/sdaa2' on '/mnt/disks/'. Oct 4 16:01:08 Tower unassigned.devices: Issue spin down timer for device '/dev/sdaa'. ### [PREVIOUS LINE REPEATED 1 TIMES] ### Oct 4 16:04:20 Tower unassigned.devices: Unmounting disk ''... Oct 4 16:04:20 Tower unassigned.devices: Cannot unmount '/dev/sdaa2'. UD did not mount the device. Oct 4 16:04:20 Tower unassigned.devices: Disk '' could not be unmounted. Oct 4 16:04:31 Tower unassigned.devices: Issue spin down timer for device '/dev/sdaa'. Oct 4 16:04:32 Tower unassigned.devices: Unmounting disk ''... Oct 4 16:04:32 Tower unassigned.devices: Cannot unmount '/dev/sdaa2'. UD did not mount the device. Oct 4 16:04:32 Tower unassigned.devices: Disk '' could not be unmounted.

It will move all the data on that share to the array, it won't leave any symlinks, files can't be in use, e.g., for VMs they must be shutdown.

You can use SSDs on the array, I have a small one I've been using for may months, like mentioned trim won't work so write performance might suffer some with time, also recommend having a faster device (like an NVMe device for example) for parity (also with higher endurance) since parity will have much more writes than the array devices, the more array devices you have the more parity will suffer in comparison, and parity can never be trimmed even if/when Unraid supports array device trim in the future.

When you start the array manually (with array auto-start disabled) VMs won't start automatically.

To make the pool not degraded. These are only one operation, balance is what converts the pools to single mode.

Set it to raid, that's Intel fake raid, and it should make it use the AHCI driver. You should also run an extended SMART test on parity, since SMART is showing some issues.

Does it do the same if you boot in safe mode?

Your RAM is 2666Mhz, and that's supported by the CPU, that's not the problem, unless the RAM is bad.

This should help with Windows VMs: https://forums.unraid.net/topic/51703-vm-faq/?do=findComment&comment=557606

Run always non correcting checks from now on (unless there's an unclean shutdown), if one finds more errors in the future immediately run another non correcting check without rebooting and post those diags.

If any corruption is detected in a redundant btrfs member it's automatically corrected on read (also during a scrub).

Not yet, but a plugin was released for that recently:

Disable array auto start, then you can disable VM autostart

SMART looks mostly OK, since you already swapped cables just monitor, if it happens again to the same disk it's likely a disk problem.

Not likely, like mentioned above it's quite common with Ryzen boards during heavy i/o, suggest upgrading to beta before trying to rebuild.

If that's an option disable autofan logging, also cache got full a couple of times resulting in the docker image failing to write: Oct 1 20:55:30 Zeus shfs: cache disk full ### [PREVIOUS LINE REPEATED 71 TIMES] ### Oct 1 21:07:50 Zeus kernel: loop: Write error at byte offset 3345833984, length 4096. Oct 1 21:07:50 Zeus kernel: print_req_error: I/O error, dev loop2, sector 6534832

Note that if you uncheck all disks like I mentioned they all will be used for user shares, including any future ones, so no need to remember it.

Please use the existing plugin support thread:

https://wiki.unraid.net/Troubleshooting#Re-enable_the_drive

Looks fine, most likely a connection issue.

Once a disk gets disabled it needs to be rebuilt/resynched: https://wiki.unraid.net/Troubleshooting#Re-enable_the_drive

That disk has SMART disable, when a disk is part of the array Unraid enables it automatically, since it's currently unassigned you need to enable manually: smartctl -s on /dev/sdX Then grab a report.

Yes.

Once a drive gets disable it must be rebuilt/re-synced: https://wiki.unraid.net/Troubleshooting#Re-enable_the_drive Make sure the emulated data drive is mounting and contents look correct before rebuilding on top.