JonathanM

Unless you set a new config and change the logical layout of the drives, no need to rebuild parity. It can stay valid while you reformat the disks and copy data back to them. Parity doesn't hold file data, the entire formatted file system is all part of parity, so reformatting the drives doesn't upset parity, it can stay in sync the whole time.

Yes.

This sums up probably 95% of the me too posts in this whole forum, not just this thread.

The OS itself, yes, that is quite easy. Simply download the manual installation zip file and extract the bz* and changes.txt to the root of your registered flash.

It will work that way because you are writing specifically to /mnt/plex-meta-data/Library/appdata/plex/, but if you or an app writes to /mnt/user/Library/appdata/plex/ it will go on the array and stay there. Turn on help and read the description for "Use cache pool". You really need to work through how this stuff works so you can understand what went wrong originally, and how Unraid decides where files written to user shares are sent, and how user shares and disks / pools interact. I recommend watching Spaceinvader One's youtube on Unraid shares.

Go to the shares tab, click on the Library share, and verify the settings are correct.

Make sure your Library share is set to keep the files only on the plex-meta-data pool, otherwise the mover will pull the rug out from under you.

Depends. I've seen several instances where they broke, but all of my normal single password instances are still running (for now). I assume it will catch everyone eventually.

Parity has no format or file system.

*.key

Could we get an option in the config like there is for access mode?

did you copy your license file?

Are you saving and applying the new config before you leave the page?

docker exec <container name> /full/command/that/works/at/the/container/console

When the CPU is reporting high temps, is the heatsink too hot to touch? If yes, airflow is the issue. if not, heatsink compound / mating surface flatness / mounting tension is the issue.

6 Bytes is probably not consequential, but if you click on the view button to the right of Disk1 on the main GUI page, you can browse what exactly is in the appdata folder and make a decision on whether it seems important or not. Short answer, stop the plex container, move all the files and folders that are currently in /mnt/plex-meta-data into a newly created /mnt/plex-meta-data/appdata/plex/ folder, change the host mapping on the plex container, start it back up. There are many different ways to move data around on the server, are you familiar / comfortable with one in particular? mc at the console would be my pick, there is a newly launched plugin that adds file management into the main GUI view buttons, and there are ways to move data directly over network shares as well, but that could possibly cause other issues.

Yep, /mnt/plex-meta-data/appdata/plex/ should work fine. Before you change it though, you should probably go to the shares page, and click compute on the appdata share. Apparently there is some data from that share on disk1, it should probably be moved to the cache pool if there is room. You should look on disk1 in the appdata folder to see what's there.

Quickest way would be to change /mnt/plex-meta-data/ to /mnt/plex-meta-data/appdata/plex/ , assuming your appdata share is not set to move or overflow to the array. Attach diagnostics if you want better advice.

The names shown indicate that the offending container is one published by linuxserver.io. Look for a host mapping showing /mnt/<disk or pool> instead of /mnt/user

You may be able to manually update by extracting the new files onto the flash drive manually, but I wouldn't count on being able to use only 2GB for very much longer.

FTFY

I agree this is needed, but I suggest not relying on the host VNC for daily use, rather set up something like nomachine in the client, and relegate the host VNC to situations where you must have baremetal type management.

ECC memory should issue a code that shows up in the logs, or if it's bad enough it should hard lock the machine to keep from corrupting more data.

@AndrewZ @SpencerJ