itimpi

Unraid 7.2 IS using the current LTS Linux kernel, so you would need a custom kernel if you cannot wait for the next LTS kernel (which as far as I know currently has no predicted date).

If you do not get the issue in Safe Mode then it strongly suggests that a plugin is causing the issue.

In my experience if anything it is safer than doing the GUI update.

Yes you could recover, but the chances are the same whether running in normal or maintenance mode.

Since there is nothing on the 4TB drive you want removed then using New Config gives you a much faster option with minimal downtime. Start with New Config and use the option to keep current assignments Change the parity drive to be the new 26 TGB one you want to use, and change the 4TB drive you want to remove to be the old 14TB parity drive Start the array to commit the new assignments and begin rebuilding parity. Unraid will not touch the data on the existing data drives. Note that until the parity rebuild completes you are not protected against another drive failing. The array is available during the parity rebuild although performance for accessing the data drives is degraded. Format the new 20TB drive to make it ready for use. This can be done while parity is being built if you want.

That was true in earlier Unraid releases but it has not been the case for some time now and the clear runs as a background task having little impact on system performance. What is true is that the drive is not usable for storing data until the Clear finishes and you format the drive.

You are correct that parity2 uses a different calculation to parity1, but there is no problem having parity2 and no parity1. Never heard before about parity2 causing any sort of stability issues.

Your diagnostics shows appdata shareUseCache="prefer" # Share exists on cache, disk7 system shareUseCache="yes" # Share exists on disk6, disk7Both of these will impact performance as you want them entirely on the cache for best performance. To rectify this: Disable the Docker and VM services under Settings Change the 'system' share so that it has the mover direction as array->cache Run the mover to get the content of these shares moved to the cache Once the move has completed you can remove the Secondary storage location for both shares so that they become "cache only" shares Go to Settings->Global Share settings and enable the Exclusive shares option. This helps performance by by-passing the Unraid Fuse layer for cache only shares. Stop and then restart the array to activate the Exclusive share option for the 'appdata' and 'system' shares. You can now re-enable the docker and VM services. You should now have Docker optimised for best performance.

You are likely to get more informed feedback if you attach your system's diagnostics zip file to your next post in this thread. It is always a good idea to do this to allow us to see the current state of your system and so we can see logs. Did you disable the docker service before running mover to get files moved back to the cache? This is to make sure that files are not being kept open as that will stop them being moved.

Since you disabled CSM, did you check that the EFI folder on the flash drive does not have a trailing "-" character? If so remove it so that UEFI boot is enabled.

No idea why a share should no longer show up - that is not normal. You are likely to get more informed feedback if you attach your system's diagnostics zip file to your next post in this thread and mention the name of the problem share. It is always a good idea to include diagnostics to allow us to see the current state of your system and so we can see logs.

Your approach would work. If you do not mind ending up with a parity2 and no parity1 then steps 2-6 could become. add old 20TB parity1 disk to array AND new 28TB disk. Start array and let Unraid "Clear" them (which is faster than preclear). Note that preclear is really intended to stress test a drive so you can instead let Unraid "Clear" drives as needed if you do not want the stress test.

You would need to somehow add reiserfs support. Not sure if that is available if you enable WSL2 in Windows since it is a standard Linux file system.

Good question. The Help text says it will always start Enabled on boot and this cannot be changed. Sounds a bit weird to me as I would have thought for security reasons it should default to disabled. Also very strange that the settings from the FTP server settings are not saved for subsequent boots..

It is going to depend on when the next LTS Linux kernel is released (which as far as I know does not have a date) as Unraid only uses LTS kernel releases.

Check in the config/shares folder on the flash drive that the capitalization of the .cfg file for that share matches the folders for that share on your drives.

Your syslog is full of messages like the following: Oct 28 11:55:14 Tower mcelog: Fallback Socket memory error count 16466 exceeded threshold: 29058440 in 24h Oct 28 11:55:14 Tower mcelog: Location SOCKET:0 CHANNEL:? DIMM:? [] Oct 28 11:55:14 Tower mcelog: Running trigger `socket-memory-error-trigger' (reporter: sockdb_fallback) Oct 28 11:55:14 Tower mcelog: Corrected memory errors on page fffa15000 exceed threshold 22 in 24h: 22 in 24h Oct 28 11:55:14 Tower mcelog: Location SOCKET:0 CHANNEL:0 DIMM:? [] Oct 28 11:55:14 Tower mcelog: Fallback Socket memory error count 935 exceeded threshold: 29059376 in 24h Oct 28 11:55:14 Tower mcelog: Location SOCKET:0 CHANNEL:? DIMM:? [] Oct 28 11:55:14 Tower mcelog: Running trigger `socket-memory-error-trigger' (reporter: sockdb_fallback) Oct 28 11:55:14 Tower mcelog: Fallback Socket memory error count 935 exceeded threshold: 29059376 in 24h Oct 28 11:55:14 Tower mcelog: Fallback Socket memory error count 16466 exceeded threshold: 29058440 in 24h Oct 28 11:55:14 Tower mcelog: Location: SOCKET:0 CHANNEL:? DIMM:? [] ### [PREVIOUS LINE REPEATED 1 TIMES] ###Not sure what they mean but it looks like something that is RAM related.

Make sure that you have a copy of the licence key file from the current flash drive (it would not hurt to have a backup of the entire 'config' folder in case you want to keep/check any settings from the current setup). You then redo the creation of the flash drive as you would for a new system and put the licence file into the 'config' folder on the flash drive.

If these are existing files, then they are not automatically moved to the location specified. If these are new files then something is configured incorrectly to make this happen.

Not quite. I am saying that all top level folders on any drive will show up as User Shares. That does not stop you using a physical path - just expect any particular file to show up under both paths as they are just different views. As was mentioned at one point it was recommended to use a physical path to get maximum performance by by-passing the Fuse layer in Unraid. Nowadays using an Exclusive share also by-passes Fuse and gets the same performance as the physical path.

The /mnt/user/sharename type path is a logical consolidated view that will show the files in that share regardless of where they are located (i.e. on the array or any pool). Files will also appear under the physical device they are located on if looking directly there.

This suggests that the system may be crashing rather than trying to shutdown and it being unable to stop the array (thus leading to an unclean shutdown). In term of it restarting you should check that you do not have the BIOS set to automatically start the server if power is present.

That is not unexpected when writing to the parity protected array due to the way Unraid handles writes and the overhead of updating the parity drive. That is one reason the concept of a ‘cache’ drive was introduced so the initial write could happen at device speed, and the slower transfer to the array happened overnight when the system was typically idle. If speed is a key criteria then you want to be using pools rather than the main array. This need was one of the main reasons ZFS support was added to Unraid.

This would not happen as the Split level precludes that unless you are putting them into the ‘2160p Comedy’ or ‘2160p Action’ which already exist on disk4. You would manually need to create the ‘2160p’ folder on disk4 to get it start being used. In addition the allocation method is probably not relevant in your case as split Level overrides allocation method. However for reference High Water is normally preferred as it minimises switching between drives.

No. You need to let the parity swap process complete (I.e. the rebuild) before making any other change to the array