Squid

Unraid has no "power saving" on the CPU unless you install the sleep plugin (which you don't have)
 
It would be worthwhile to not run your memory at an overclock (3000MT/s)  The memory you purchased F4-3000C15-8GVR is actually 2133.  Disabled XMP / AMP is the BIOS and also run a memtest from the boot menu for a minimum of a pass - If you're booting the OS via UEFI, you will need to temporarily switch to Legacy boot in order to run the test

Fix Common Problems - Tools, Update Assistant

Something happened on your flash drive.  Delete /config/plugins/community.applications.plg and /config/plugins/community.applications
 
Reboot and you'll be good to go after a reinstall

Yeah, that is going to be your entire problem.  Both the logs and CA's diagnostics show that the date is March 15, 2019 and SSL fails if the date is significantly out for security reasons.
 
Settings - Date & Time. set the correct date & time and also give the system a couple of ntp servers
 

AMD Ryzen 9 5950X 16-Core Processor  
Have you looked at this yet:  https://forums.unraid.net/topic/46802-faq-for-unraid-v6/page/2/#comment-819173
 
Specifically Typical Idle Power Supply Settings in the BIOS?
 
Also, for stability you should never ever run memory at an overclock.  You're running your memory with the XMP profile at 3200MT/s
 
It *appears* that you bought "generic" memory, since the Manufacturer and product number is "Unknown", but since 3200MT/s memory is exceedingly rare, you're more than likely running 2133, and should disable XMP in the BIOS.

Then you've done a symlink or something because the situation with emby is not actually possible. 

Click on each share (shares tab) and you can set the minimum free space.  It should however actually default to 0

Have you run Memtest from the boot menu at least a pass or two?  (If you're booting via UEFI, you will need to temporarily switch to legacy for Memtest to work)

Jul 6 23:12:49 Axiom mcelog: failed to prefill DIMM database from DMI data  
Primarily an informational warning, and safe to ignore.  A BIOS update might help though.

Is your reverse proxy allowing websockets through?  (It's usually an option)

Not necessarily your problem, but try running your memory at its rated speed of 2133 instead of it's overclocked XMP profile of 2666
 
In addition, run a Memtest from the boot menu

Segfaults are rather common with Plex.  You could run a memtest to rule out any memory issues, but the best place to complain about it is on Plex's forum

The above still applies - Within same share, the file stays on the same disk.  Different shares the destination will follow the use cache rules on the destination.  Using a rootshare file stays on the same disk

Yes
 
Via Dynamix File Manager (or alternatively Krusader).
 
If you have a parity drive, then under Settings - Disk Settings turn on Write Method: Reconstruct Write

All of the "games" in Apps aren't games, rather Game Servers (as the Category says).  You're not downloading the game (which you have to buy on Steam), but rather a game server which allows you and your friends to connect to to play in your own separate world separated from the rest of the players online (along with other things)

It would appear that the mce was issued during the microcode update, and is more than likely nothing to worry about at all and safe to ignore.
 
Upgrading the BIOS however from the stock pre-release version you're currently running to the release version may also be helpful.

The docker console isn't supposed to show any logs.  Rather it's you logging into the command prompt for the app.  What you're seeing is exactly what is supposed to show up.
 
Are you meaning instead to be selecting the "Logs" item in the dropdown?

Clicking anywhere on the card (or the info button) brings up the sidebar which will tell you that it requires Unassigned Devices, and that requirement isn't met (hovering over the + on the card will also tell you that it requires UD installed)

Because part of the process in rebooting or powering off is stopping the container.  Therefore it doesn't meet the criteria of "restart unless stopped", and since it's via docker compose it's outside of the control of the OS' GUI and the autostart flag.

Those 3 files are definitely corrupted.  Delete them

May 26 11:01:06 Tower mcelog: failed to prefill DIMM database from DMI data May 26 11:01:06 Tower mcelog: Kernel does not support page offline interface  
More than likely not an issue and safe to ignore, but maybe a BIOS update would solve it

Apps, previous apps, reinstall jelly fin and fix the error.
 
Not an OS issue because due to the failed docker run command the app is no longer actually installed.  Prev apps gets the template back the way it was last saved

What are these lines in your diagnostics about?
May 27 14:56:19 Media webGUI: Unsuccessful login user root from 100.81.154.14 May 27 14:56:24 Media webGUI: Successful login user root from 100.81.154.14 Not necessarily a problem, but 100.x.x.x. is usually reserved for CGNAT and isn't something you see every day.
 
You are also having disconnects on the drive(s).  Reseat the cabling.  In extreme cases, this can cause what you're seeing.

They look to be from a container continually restarting itself.  Start by looking at the uptime of each and go from there.

Download the zip file from the website, then overwrite all of the bz* files on the flash drive with those in the zip.