So, long story. I've got an HP DL380 G9. I've had my 2 SSD's just sitting in 2 bottom drive trays since I didn't have an SSD drive adapter. (The server is outfitted for 3.5" drives) Well, I purchased adapters and they just came in today. I stopped the array, popped out the SSD's, put them into the adapters, and then put them into the server. Ended up having to use different slots, since the adapters didn't want to fit right for some reason... Anyway. Started the array back up and now it's saying that the cache is unmountable.

If I have to start format the cache, it's not 100% the end of the world. My appdata backs up every night to the array. But I do have some shares that use the cache that files got copied to today.

 

When starting the array in maintenance mode and checking the file system on the cache drive I get this output:

 

No valid Btrfs found on /dev/sdk1
ERROR: cannot open file system
Opening filesystem to check...

Are there any steps I should take before nuking the cache? Thanks in advance!

 

//edit//

 

Reading some other threads I've been throwing some commands at the server.  Ran btrfs check -s 1 /dev/sdk1 and got:

 

using SB copy 1, bytenr 67108864
Opening filesystem to check...
bad tree block 298118643712, bytenr mismatch, want=298118643712, have=0
Couldn't read tree root
ERROR: cannot open file system

 

Edited January 26, 20206 yr by my95z34

Go to Tools - Diagnostics and attach the complete diagnostics zip file to your NEXT post.

I never got a notification for that comment.  Ugh, lol.  I ended up just nuking the drive.  I remembered that Syncthing had backed up everything that I thought I'd lose, to my other server.  Plus, I run CA Appdata backup nightly, so I only lost a small amount of Plex watched statuses.

 

I can attach it now, but I doubt that'd do any good.

 

Also, just for clarification, was removing drives while the array was stopped, but leaving the server powered up, a bad practice?