trurl

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

Your syslog is full of errors related to your Unassigned Devices. Not sure what that is about but I am guessing you have NTFS filesystem corruption on one of them. You might try plugging them in to your Windows computer and running checkdisk on them.

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

If you had posted Diagnostics we could have examined SMART for all disks.

You have to share it in Unassigned Devices.

Split level would be meaningless if it didn't.

Did they provide the requested information? You should zip and post.

Also, how exactly are you copying files to your server?

You should set Minimum to larger than the largest file you expect to write to the share. When Unraid chooses a disk to write, it has no way to know in advance how large the file will become. If a disk has less than Minimum, it will choose another disk. Krusader, rsync, and some other applications write all the folders in advance. Since a bunch of empty folders likely will not reach Highwater, all the folders get written to the same disk then the files get written to those folders. Post your diagnostics.

Have you tried booting in SAFE mode? Technically you should test in SAFE mode before reporting a bug, and I see you have some very old plugins installed, plugins that have been deprecated for years and should have been replaced long ago by dockers. Nzbget.plg - 2016.09.17.1 Sonarr.plg - 2017.02.15.1 Transmission.plg - 2016.09.17.1

I see some Most-Free shares and some High-water shares. Is this the way you have always used them, or did you make changes? Which shares are you writing that is continuing to go to disk1? If you have continued to write to your server and/or made changes to any settings since those diagnostics you should post new ones.

Do you know how to get to the folder from the Unraid command line and list the contents with their permissions?

Sorry my bad, not your controller, it was the fact you had SAS disks instead of SATA that had me confused.

Since we never got any reply from OP I guess you can have this thread, if you

No parity doesn't have any data. There is no magic that would allow parity to have the capacity to be a backup of all the other disks. Parity is just an extra bit that allows a missing bit to be calculated from all the other bits. And Unraid parity is real time, so as soon as you deleted your data, parity was updated and agrees that the data is deleted.

OK, I did briefly look at your diagnostics. You seem to be using a controller that doesn't pass the disk serial number to Unraid. So I don't know if your disks will be identified exactly as before when you remove disks. You may indeed have to reassign everything. Have you had problems with this in the past?

A parity sync (rebuild), which is what you need it to do, so the removed drives are no longer part of parity. No reason at all to do anything with your existing cache. Tools - New Config, preserve all assignments. Then it will let you make any changes you need before starting the array and rebuilding parity. Just unassign the disks you want to remove and start the array. Be sure to NOT check to box saying parity is valid. I didn't actually look at your diagnostics. Is there something there you may be concerned about?

https://forums.unraid.net/topic/61211-plugin-ca-appdata-backup-restore-v2/

User shares are not cached by default. Unless you set a user share to use cache, it will not and its files will be written directly to the array. Almost all of my user shares are not cached. You can also have user shares that stay on cache and aren't moved to the array. This is often done when performance is important. Typically appdata (docker working storage), domains (VMs), and system (docker executables and VM configurations) shares are set to stay on cache so dockers/VMs will perform better and so these will not keep array disks spinning.

You need an external (your router address on the internet) static IP or else dynamic DNS, duckdns is often used for that. Getting the external IP forwarded to ports of the internal IP of your server is what port forwarding is all about. Have a look at SpaceInvader One youtube channel: https://www.youtube.com/spaceinvaderone

That is not the diagnostics, just the syslog. See if you have a diagnostics zip file.

User shares are how Unraid allows folders to span disks. When you create a user share, Unraid creates a top level folder named for the share on cache or array as needed in accordance with the settings for the user share. Conversely, any top level folder created on cache or array is part of a user share named for the folder. If you don't make settings for a share it has default settings. Cache is part of the user shares, and each array disk is part of user shares. If a top level folder exists on cache, and a top level folder with the same name exists on an array disk, then those folders are part of the same user share. You might try deleting the appdata for plex in addition to deleting the docker and reinstalling plex from scratch, but that would mean you would be starting over with plex and would have to rescan all your media.

I always let my router handle IP assignment and let all my computers and devices use DHCP. That way everything can be managed in one place. For my wired devices, I reserve IP by MAC address. If you want to move everything to a different subnet than the router is already configured for you will have to reconfigure your router. I would suggest not doing anything until you have a better idea of what to do. Random experiments will make things inaccessible, and possibly open them up to hacking.

SMART for parity looks OK. To rebuild to same disk: Stop array Unassign disk Start array Stop array Reassign disk Start array to begin rebuild

Do you have Notifications setup to alert you immediately by email or other agent as soon as a problem is detected?