itimpi

Sounds as if you have misunderstood how Unraid handles User Shares. The cache is NOT a copy of files on the array. Files in Unraid exist either on a pool OR on the main array - not at both locations. User Shares are then a unified view of files covering both the array and any pools.

Have you tried looking at the source code for the Unraid md driver as it is included with Unraid. I would not be surprised if the calculations are in there somewhere (although I admit I have not looked)

Not quite sure what the confusion is? Parity1 can calculate the correct bits for 1 failed disk. Parity2 can do the same for a second drive using a different algorithm (ignoring parity1). As such they are independent of each other. if you want the exact details of the parity2 algorithm then I have no idea other than the fact it includes the drive slot as one of the elements and that the algorithm is such that a 0 byte on a data drive does not change the calculated result

The first screenshot is the type of IP address you get when the DHCP server cannot be contacted. the second screenshot is what you get if the flash drive is not labelled UNRAID (all capitals). You might want to label it manually to be sure in case the USB creator tool did not do this correctly.

You only need to disable these services once to get the current state cleaned up for these default shares. When the files are where they should be you can re-enable them and everything should then just operate fine. The problem tends to occur when you first start these services without a cache pool attached. As a result the services support files end up getting created on an array drive. When you later add a cache pool Unraid puts any new files for the "Prefer" type shares on the cache pool and tries to move what other files it can to the cache pool. It fails to move these services support files because the services are holding them open.

Your diagnostics show that the ‘appdata’ and ‘system’ shares have files on disk1 so that could explain your symptoms. They are set to Use Cache=Prefer, but unless you disable the Docker and VM services while running mover the services will hold files open which stops mover from moving the files used by theses services to the cache.

You are likely to get better informed feedback if you attach your system’s diagnostics zip file to your next post in this thread so we can check how things are set up.

It is not clear to me what is not working! Are you saying that new downloads that should be going (at the qBitTorrent level) to /Movies are going somewhere else other than disk5? If so where are they now going?

Then I am confused as your earlier post said you got the key in the email Did you actually get a download link instead?

MacOS is a case sensitive OS so a Yes value seems appropriate (and is also more efficient at the Samba level).

If you got the key file attached to an email then it should just be a case of putting that file into the ‘config’ folder on the flash drive. Not sure from what you said if you tried that?

Unraid is largely hardware agnostic. It should just be a case of plugging everything into the replacement motherboard and booting it. Unraid does not care where drives are connected - it recognises them by their serial number. The one thing that can cause some issues if you are using RAID controllers on one system and an HBA on the other so drives serial numbers are not reported in the same way.

Losing settings implies problems with the flash drive as that is where Unraid stores all settings in the config folder.

No - a GPU cannot be used by a docker container and a VM simultaneously.

The minimum RAM requirement for current Unraid releases is now 4GB.

You should remove all these until after you get things working as these can cause problems if Windows tries to connect them with the wrong user.

You can already add extra arrays as btrfs pools, with ZFS pools coming shortly as an alternative in the 6.12 release. Multiple arrays of the Unraid main array style is still a roadmap item but no ETA is even hinted at so far as to when that might be svailable.

Yes, because the parity drive has no file system.

I am afraid I do not use Docker Compose so not able to help. Hopefully someone else who is experienced in Compose can help.

If mover has a lot of very small files to move then they tend to take MUCH longer than the same total size in larger files.

The memtest included with Unraid only runs if you are booting in Legacy/CSM mode. If you want a version that will run with UEFI boot you can get one from memtest86.com

What was/is the output from the xfs_repair? I would expect it to give an indication of why it is not working.

It could be that it only went read-only when attempting to write the new plugin files to it.

You should be able to run the same container on Unraid. if the container is started using Docker Run then as it appears nobody has created an Unraid specific template it is just a case of providing the relevant parameters in the blank template Unraid provides. if the container is started via Docker Compose then there an Unraid plugin for Docker Compose to help with doing it that way.

It is likely to help get a diagnosis if you can get a system's diagnostics zip file when the problem occurs and add that to your next post in the thread.