JorgeB

Can you ping, for example google.com Also, please post the diagnostics

Please post the diagnostics, and ping me again after that, or I may miss the post since it's not in the English section.

Yep, the reports go directly to Squid.

Reported ATA issues are with parity and parity2. Check/replace cables for both disks, could also be a shared power splitter if one is in use.

Please post the diagnostics.

There are ATA errors in two drives; are there any power splitters in use? Also, please post the full diagnostics.

Not really, since I never used Plex; someone else may be able to help, but the best place is the respective container support thread. You can also try to remove it and reinstall it from previous apps in case it's an image issue.

If now the new NIC is eth0 you can disable bonding. This will also work, but IIRC bonding two different speed NICs limits the speed to the slowest one (if both are connected).

I would still recommend changing that, since overclocking a server is never a good idea, and with those CPUs, it's known to sometimes cause data corruption, like errors during a parity check.

If you leave the bond enabled, it can use eth1, if you prefer to disable it, then you need eth0 to be the main one.

It downloaded fully, but since you hadn't selected to keep the file in Chrome, I had to manually rename it to zip, but the diags are complete.

Docker service is starting correctly as far as I can see, if the problem is only Plex, best bet is to look for its support thread:

I understood you manually changed the MAC address on the network-rules.cfg file, if you used the GUI the change was not done correctly or it didn't take; the new NIC is still eth1, try doing it again, reboot, then post new diags.

This means the key was not valid. Possibly it caught the new key in the process of being written. Since the paid key was invalid, it looked for another one and found an expired trial, but this by itself is not a problem, since it would not matter if a valid key was installed. It then detected the correct new key, but the array had already started to stop. The array failing to stop is unrelated to the key issues and likely related to this, which suggests something is writing something to appdata outside Unraid. This is the first time I ever seen this happening, so possibly an isolated/timing issue, but I'll test myself and will forward this issue to LT even if I can't repro in case this can be improved.

The kink posted mentions the max speed officially supported by AMD depending on the CPU and RAM used; anything above that is overclocking.