trurl

Go to Tools - Diagnostics and attach the complete Diagnostics zip file to your NEXT post.

Simpler to use Previous Apps on the Apps page. We never got your Diagnostics. Those might give us a better idea of what went wrong. Possibly you still haven't really fixed the cause of the issue and it will return. Go to Tools - Diagnostics and attach the complete Diagnostics zip file to your NEXT post.

The first syslog you posted suggests a flash problem. Are you booting from a USB2 port? (recommended) Have you done memtest?

You have quite a few warnings and errors from Fix Common Problems. Why haven't you done anything about them? Not quite understanding this part: If it works in SAFE mode then that means one of your plugins is to blame. Don't know why you would say "of course everything gets wiped out if I reboot." Not obvious to me that anything would get wiped out if you reboot. Could you elaborate on that? One thing I don't understand in your diagnostics, you have an old version of the Nvidia plugin, but it doesn't appear you are using it since your Unraid version is much newer than that.

Are you sure you didn't copy these files to Unraid from some other system? My guess is they are the result of filesystem repair but I don't remember seeing these particular examples from a repair on Unraid.

So where exactly does preclear fit into this narrative?

Preclear doesn't format a disk, and neither does UD unless you tell it to. After a preclear a disk can't be XFS or any other filesystem

That is a screenshot from the Dashboard

What said it was XFS? Unassigned Devices?

I'll just see if @bonienl can make any sense of this.

Feb 26 04:40:01 Tower root: Fix Common Problems: Error: Same share (html) exists in a different case Feb 26 04:40:01 Tower root: Fix Common Problems: Error: Same share (HTML) exists in a different case The lower case version doesn't appear in the diagnostics, but that may just be an artifact of the symlink. The correct fix for this would be to only have one of these folders with all of their contents combined, and to figure out what you did to create the other one. Accidentally creating another user share with different case is a common result of using incorrect case in a docker mapping.

I don't think Unraid will clear a disk unless there is parity. There certainly is no reason for it to do so since the whole point of clearing a disk is so parity will remain valid.

Go to Tools - Diagnostics and attach the complete Diagnostics zip file to your NEXT post.

Not according to your diagnostics. You have disks installed and assigned, but none of them are mounted. Have you let Unraid format them yet? Post a screenshot of Main - Array Operations.

Go to Tools - Diagnostics and attach the complete diagnostics zip file to your NEXT post. Diagnostics contains syslog and many other things which give a more complete understanding of your situation.

So did you ever setup a parity drive? Some of the discussion after that really only applies if you have a parity drive. For example, your very next post, and the response it got: None of that makes any sense unless you do actually have parity. Also, are you still trying to use USB for any of your array (or cache) disks? The unreliability of USB connections pretty much defeats the purpose of having parity. In fact, USB in an array with parity will likely result in constant rebuilds.

Well of course those paths don't work from the Unraid command line. I guess you must be thinking of making a round trip to another computer on your network using disk shares. And I don't recommend sharing disks anyway. Do you have Krusader installed? I would just use Midnight Commander (mc from the command line) since it is builtin. I never work with disk shares over the network, and only work with the disks from mc or Krusader.

How do you plan to do the move? Command line or some other method?

Many people make the mistake of not realizing that the drive "letter" isn't constant from one boot to another, so they can easily get /dev/sd? wrong by not checking which disk a particular identifier refers to. The drive letter is especially likely to change if you plug or unplug any disks. Another thing that people who know about the command line might do is try to work directly with the sd device, when they really need to work with the md device. If you work directly with an sd device that is part of the parity array, you will invalidate parity.

Just thought I would add another cautionary note to all this. Another way people have lost data in this situation is by using a flash backup that isn't current. Here is how this happens. Sometime after making a flash backup, you assign a new parity disk. Then, you add that old parity disk as a new data disk. Later, you need to use the flash backup, but it still thinks that new data disk is parity. And you have the array set to autostart. So, Unraid starts writing parity to that original parity (new data) disk, overwriting its data. You should get a flash backup any time you change any settings you don't want to lose. And you absolutely must get a new flash backup if you make any changes to disk assignments.

If there are any writes of any kind from anything to the emulated disk then it is definitely going to need to be rebuilt since the pulled disk will definitely be out--of-sync with parity.

Go to Tools - Diagnostics and attach the complete diagnostics zip file to your NEXT post. Since the OP hasn't visited this thread for a while I guess you can have it.

I have split your post into its own topic so we can leave that other thread for supporting that other user. Go to Tools - Diagnostics and attach the complete diagnostics zip file to your NEXT post.

This is usually spelled "appdata". If you actually used "Apdata" then you may have created some problems.

Are you sure you don't have some docker mapping with an incorrect Host path? Any path that isn't actual storage is RAM.