JorgeB

No need to flash, it's IT mode only.

Your cache drive is completely full, this is preventing writes on the libvirt loop image, free up some space, re-start the VM service and it should return to normal, but it's always a good idea to backup libvirt.img, in case it gets corrupted.

If it was an unclean shutdown you're fine, and it's the most likely, if it was a disk returning bad data not so good since there would be some data corruption, but like mentioned that is pretty rare.

It is very strange to now be getting an "invalid partition error", this means something changed the mbr/partition and it does no longer conform to what Unraid expects, that should never happen (unlike filesystem corruption which can happen sometimes), I suggest you reformat disk1 (any data there will be lost), copy some data then stop and re-start the array, if it goes unmountable again post new diags before rebooting.

There must be a reason for the errors, most obvious would be an unclean shutdown some time before this check, another possibility is one of the disks returning corrupted data, that's not supposed to happen but it is known to happen on some rare occasions.

I would, the pending sectors are real since it failed the SMART test, also these attributes look very bad: They should be zero, or at least very low, on a healthy WD drive.

Please post the diagnostics: Tools -> Diagnostics

Most likely, either those are new disks or something else happened that wiped the partitions, if the latter a file recovery util like UFS explorer should still find the data.

You need to format the disks, next to array start/stop buttons.

Yes

Do you really have 1TB of RAM? I'm guessing that's confusing Unraid, probably first user with so much RAM.

2.5GT/s means a PCIe 1.0 slot, if it was also a x4 link instead of x8 it would only have a quarter of normal bandwidth, so big impact is expected.

Diags don't show any disk1 issues, they do show read errors on disk4 (looks more like a power/connection problem) during the pariuty sync, so parity is invalid, because of the log spam caused by those errors it doesn't show anything more after that. Please post current diags.

New config didn't take, all disks must change to a blue icon, repeat it. You missed the "Apply"

Mover won't overwrite any existing files, you can enable mover logging and see why they are not being moved, or post the diags after doing that.

The "invalid partition error" suggests those disks weren't array data disks, or something happened to them and the partition/mbr was changed and is not what Unraid expects, unassigned them all and check with the UD plugin if a filesystem is detected on any of them, UD won't care for the partition layout, as long as there is a partition.

Yes, but always from disk to disk, never disk to share or share to disk.

This started with the disk dropping offline: Mar 4 08:24:11 Tower kernel: ata4: link is slow to respond, please be patient (ready=0) Mar 4 08:24:15 Tower kernel: ata4: COMRESET failed (errno=-16) Mar 4 08:24:21 Tower kernel: ata4: link is slow to respond, please be patient (ready=0) Mar 4 08:24:25 Tower kernel: ata4: COMRESET failed (errno=-16) Mar 4 08:24:31 Tower kernel: ata4: link is slow to respond, please be patient (ready=0) Mar 4 08:25:00 Tower kernel: ata4: COMRESET failed (errno=-16) Mar 4 08:25:00 Tower kernel: ata4: limiting SATA link speed to 3.0 Gbps Mar 4 08:25:05 Tower kernel: ata4: COMRESET failed (errno=-16) Mar 4 08:25:05 Tower kernel: ata4: reset failed, giving up Mar 4 08:25:05 Tower kernel: ata4.00: disabled Then it came back online, before dropping again, so it suggests a power/connection issue, especially if SMART is OK, replace/swap both cables or slot with a different disk to rule that out, if it happens again to the same disk after that it could be a disk problem.

Check the board bios, usually there's more than one virtualization option, vt-d might be enable but not vt-x, also look for a bios update

Depends if the controller will be installed on a CPU or PCH slot, if PCH it will share the bandwidth with the onboard SATA ports, there are also some performance numbers for that in the link above.

Correct, but 2GB/s is more than enough for 4 disks, 4 SSDs would be a different story.

^this

Those are the same, still recommend changing the paths to /mnt/cache instead of /mnt/user, no idea what happened to the dockers since there weren't any errors I could see, but that image looks brand new based on the transid, you can quickly add them back using the CA previous apps function.

If it was on cache it would have been re-created, but since you're using /mnt/user paths you might have two images, what's the output of: find /mnt -name docker.img

Did you re-format cache? Also please post current diags.