JorgeB

SMART test passed so the disk should be OK, try swapping cables/slot with a different disk and rebuild again, if the same the disk could be bad, despite SMART.

Dec  2 20:04:57 unraid kernel: macvlan_broadcast+0x10a/0x150 [macvlan] Dec  2 20:04:57 unraid kernel: ? _raw_spin_unlock+0x14/0x29 Dec  2 20:04:57 unraid kernel: macvlan_process_broadcast+0xbc/0x12f [macvlan]  
Macvlan call traces will usually end up crashing the server, switching to ipvlan should fix it (Settings -> Docker Settings -> Docker custom network type -> ipvlan (advanced view must be enabled, top right)), then reboot.
 

Dec 18 21:18:29 Tower kernel: macvlan_broadcast+0x10a/0x150 [macvlan] Dec 18 21:18:29 Tower kernel: ? _raw_spin_unlock+0x14/0x29 Dec 18 21:18:29 Tower kernel: macvlan_process_broadcast+0xbc/0x12f [macvlan]  
Macvlan call traces will usually end up crashing the server, switching to ipvlan should fix it (Settings -> Docker Settings -> Docker custom network type -> ipvlan (advanced view must be enabled, top right)), then reboot.
 

For the cache I would recommend backing up and re-formatting, for the disk you can check filesystem, run it without -n

This is how I do it:
 
Generate key:
ssh-keygen NOTE: when generating the key use a blank password or it will ask for it anytime you use that key.

Copying your Public Key Using SSH:
cat ~/.ssh/id_rsa.pub | ssh root@remote_host "mkdir -p ~/.ssh && cat >> ~/.ssh/authorized_keys"
 

It won't.
Jan  2 10:15:53 Tower kernel: ata3.00: HPA detected: current 5860531055, native 5860533168  

The notifications can be wrong, it's a known issue, the data on main will be correct.

Missing device was now removed, you can now change slots to 1 or add a new device to create a new mirror.

Your go file is empty, replace it with the one from a stock Unraid install and reboot, key issue may be related, if there were any custom entries there.

We have crack related entries in your go file, replace it with a stock one from the Unraid zip and reboot.

First thing:
 
Dec 26 09:40:10 Asgard kernel: macvlan_broadcast+0x10a/0x150 [macvlan] Dec 26 09:40:10 Asgard kernel: macvlan_process_broadcast+0xbc/0x12f [macvlan]  
Macvlan call traces will usually end up crashing the server, switching to ipvlan should fix it (Settings -> Docker Settings -> Docker custom network type -> ipvlan (advanced view must be enabled, top right)), then reboot.
 
Then recreate the docker image since it's corrupt, if it's was recreated recently you may have an underlying issue causing that.

That looks normal, but if it keeps repeating check to see if one or more containers are constantly re-starting, you can check based on their uptime.

Go to (Settings -> Docker Settings -> Docker custom network type -> ipvlan (advanced view must be enabled, top right)), then reboot.

Libvirt image you are using is corrupt, you will need to recreate a new one or restore a good backup.

Enable the syslog server and post that after a reboot, it will show if there was a shutdown event.
 
eth0
 
Dec 31 23:13:48 MDRONET-UNRAID1 kernel: ------------[ cut here ]------------ Dec 31 23:13:48 MDRONET-UNRAID1 kernel: netdevice: eth0: failed to initialise TXQ 42 Dec 31 23:13:48 MDRONET-UNRAID1 kernel: WARNING: CPU: 12 PID: 23349 at drivers/net/ethernet/sfc/ef10.c:2414 efx_ef10_tx_init+0xfc/0x1c0 [sfc]  

Since you didn't post the complete diags I cannot confirm, but I would guess disk1 is disabled and Unraid is trying to mount the emulated disk, which has the same original issue, if that's correct you'd need to do a new config (can check parity is valid) and start the array to format disk1.

Yes

Those errors could be old now, recreate the docker image and see if they come up again, also scrub the pool

It should always show the ones logged on parity-checks.log, unless maybe the file has some corruption.

use -L

No, you even have some unnecessary steps, you can just:
 
-shutdown and disconnect old drive, connect new one
-power on, assign new drive and start array to rebuid

You can but would recommend using Unraid pools instead of the hardware RAID, you can create a zfs mirror for example.
 
Note that a LAGG won't help with a single transfer, only if you plan to make simultaneous transfers.
 
 
You can create a new trial flash drive for the new server, once the data is copied, backup /config from old flash drive and recreate it, then copy the complete /config folder from the trial flash driver except for the key and finally restore the old key.
 
 
P.S.:
If you mean zfs on the array note that currently there's a bug affecting write speed, I hope it will be fixed soon, but something to be aware of, this bug does not affect zfs pools, only zfs on the array.

Enable the syslog server and post that if it crashes again.

So pool only has 4 members, any special reason you assigned devices 1 2 4 5 and left 3 empty?
 
Unassign all pool devices, start array, stop array, now reassign them to slots 1 2 3 4 in the same order as the zpool status, so:
 
sdg
sdh
sdj
sdi
 
Start array and the pool should import.

You can create multiple 3 way mirrors and raid 0 pools with btrfs and zfs, you can also create one 3 way mirror in the array, assign both parity disks and disk1.