Jan 11 21:22:21 NAS kernel: BTRFS: error (device sdk1) in cleanup_transaction:1942: errno=-28 No space left
Jan 11 21:22:21 NAS kernel: BTRFS info (device sdk1): forced readonly
But, the more interesting thing is that you don't actually have a cache pool named "cache" Rather you have a cache pool named cache_1tb_ssd, but was probably a semi-recent change / addition, but at the time of doing that, you left your various shares all referencing "cache" in the select cache pool option.
And the share exists on "cache" and on various disks, but since there is no pool named cache, the files all wind up in RAM. (Also you'll have to manually adjust the various docker templates you've got (show more settings), and see if anything there is directly referencing /mnt/cache/...
After you've got all that fixed up, then reboot, and we'll take it from there (with a new set of diagnostics)