JorgeB

If it's a one time thing you can ignore, if it keeps happening try limiting more the RAM for VMs and/or docker containers, the problem is usually not just about not enough RAM but more about fragmented RAM, alternatively a small swap file on disk might help, you can use the swapfile plugin:
 
https://forums.unraid.net/topic/109342-plugin-swapfile-for-691/

It sure can, and it usually will.

Post new diags after array start to see the current emulated disk status.

Pool is missing a device and has multiple profiles:
 
             Data      Data      Metadata  Metadata  System    System                             Id Path      single    RAID1     RAID1     DUP       RAID1     DUP      Unallocated Total   Slack -- --------- --------- --------- --------- --------- --------- -------- ----------- ------- -----  2 /dev/sde1 365.00GiB 262.00GiB   2.00GiB   2.00GiB  32.00MiB 64.00MiB     1.20TiB 1.82TiB     -  3 missing           -         -         -         -  32.00MiB        -     1.82TiB 1.82TiB     -  4 /dev/sdj1         - 262.00GiB   2.00GiB         -         -        -     1.56TiB 1.82TiB     -  
Depending on if you wanted this pool to be single profile or raid1 run a balance to that profile, if it fails post new diags.
 
Metadata pool has errors read/write on both devices:
 
Oct 28 03:22:03 OrbitalHub kernel: BTRFS info (device sdb1): bdev /dev/sdb1 errs: wr 73363, rd 148, flush 1448, corrupt 5299, gen 10 Oct 28 03:22:03 OrbitalHub kernel: BTRFS info (device sdb1): bdev /dev/sdc1 errs: wr 1028, rd 63, flush 23, corrupt 0, gen 0  
Run a correcting scrub and check if all errors were corrected, also see here for more info and better pool monitoring.

Since the disk got disabled it needs to be rebuilt, before rebuilding, check that the emulated disk is mounting and contents look correct, to do that, unassign disk2 and start the array.

Unfortunately there's nothing relevant logged, this usually suggests a hardware issue, first thing take a look here:
https://forums.unraid.net/topic/46802-faq-for-unraid-v6/?do=findComment&comment=819173

Reimporting the pool should fix it, stop the array, unassign all pool devices, start array, stop array, re-assign all pool devices, start array.

Don't forget that you will need to be on v6.12.8, it won't work with previous v6.12.x releases due to a bug:
 
Unassign the disk, then type:
 
sgdisk -o -a 8 -n 1:1M:0 /dev/sdX
Replace X with correct letter, re-assign the disk, if you get the same error reboot, but it's usually not needed al long as you unassign and re-assign.

There are a few drives spinning up:
 
Jan 30 21:50:05 Media-Server emhttpd: read SMART /dev/sdi Jan 30 21:50:15 Media-Server emhttpd: read SMART /dev/sdh Jan 30 21:50:21 Media-Server emhttpd: read SMART /dev/sdj Jan 30 21:50:24 Media-Server emhttpd: read SMART /dev/sdk Jan 30 21:50:32 Media-Server emhttpd: read SMART /dev/sde Jan 30 21:50:40 Media-Server emhttpd: read SMART /dev/sdf
 
 
Depending on the controller it can block all reads until they all spin up, so from 21:50:05 until approximately 20:50:45

Yes.

Cache filesystem is corrupt, suggest backing up and re-formatting.

Connect plugin is an old release, upgrade to current ( or uninstall it).

Pretty sure Unraid does not change the DNS servers, but those are a good option, and safe to use.

No, but for best performance try to get similar disks, e.g., if the current ones are 7.2k rpm get the same.
 
 

The disk is not giving a valid SMART report, if you have already tried a different port and cable it's likely a bad disk.

Device here means NVMe device(s).
 
I would start by running memtest.
 
 

Feb 15 14:41:33 voyager-unraid emhttpd: device /dev/sdk problem getting id  
This means the devices have the same ID, in this case the same serial number, this should never happen, AFAIK there’s no way to get both working, only one will work with Unraid.
 
 

Export the pool first if it's online, unassign all pool devices, start the array to reset the pool config, stop the array, reassign all pool devices in the same order as the zpool status shows, leave the filesystem set to auto, start the array and pool should import, if it doesn't post new diags.

Disks were already disabled at boot, so we can't see what happened, but SMART looks fine for both, if they pass the test, the emulated disks are mounting and contents look correct, you can rebuild on top.

Use -L

Not sure how you've manage to get the pool in that state, also did you import the pool manually?
 
You can try this, but make sure backups para up to date, in case it goes sideways:
 
zpool detach cache_sata 296863395680873687  
Then post the zpool status

Feb 15 18:22:47 Tower kernel: BTRFS warning (device sdc1): devid 3 uuid c07fdaf6-74dd-4cf3-9cf2-8c63ebccdf33 is missing Feb 15 18:22:47 Tower kernel: BTRFS info (device sdc1): bdev (efault) errs: wr 0, rd 0, flush 0, corrupt 99, gen 0 Feb 15 18:22:47 Tower kernel: BTRFS info (device sdc1): bdev /dev/sdc1 errs: wr 0, rd 0, flush 0, corrupt 637, gen 0 Feb 15 18:22:47 Tower kernel: BTRFS info (device sdc1): bdev /dev/nvme0n1p1 errs: wr 17446079, rd 16940, flush 12547, corrupt 713, gen 0  
Pool has a device missing, one device that dropped some point in the past and btrfs is detecting data corruption, I would recommend starting by running memtest, then the pool needs to be fixed.

Disks dropped offline and reconnected, this is usually a power/connection problem.

You can only add another vdev of the same type and with, so another 4 devices in this case.
 
No to both.

It's not logged as a disk issue and SMART looks OK, there are recent UDMA CRC errors so recommend replacing the SATA cable first,