I recently purchased a new 2TB NVME drive to use as a cache drive to replace my 1TB Samsung SSD that has a smart error and I think is on its way out. Last night, I decided to start the swap and followed instructions posted by user boontato here: 

After I added the NVME to the cache pool, I let things balance and then stopped the array and unassigned the Samsung SSD drive and started up the array again. I was immediately presented with the need to format the NVME drive. I stopped the array again and added the Samsung drive back to the #1 slot in the cache pool and started the array again and then both drives were listed as "Unmountable: Unsupported or no file system". Figuring I really screwed the pooch here, I decided to format the NVME and restore all my docker containers and VM's via the backups that I had. Unfortunately, there was one VM image that was located on the Samsung SSD drive that I need and can't figure out how to access now. I've tried to run btrfs rescue zero-log /dev/sdg based on some googling and someone else having a similar issue and that presented me with the following error: 

No valid Btrfs found on /dev/sdg
ERROR: could not open ctree

 

Drive also cannot be mounted via unassigned devices plugin and only shows "Format" next to it.

Any ideas as to how I can recover data from this drive? I've attached a diagnostic file. The drive in question is labeled as Samsung_SSD_860_QVO_1TB_S4PGNG0KC08521K (sdg).

apollo-diagnostics-20241212-1150.zip

Post the output from:

 

fdisk -l /dev/sdg