For some reasons my cache drives will become un-mountable after a reboot.  It seems to happen after I've been shutdown for sometime making changes on my server.  Case in point - I installed a new fan - powered up - cache drive un-mountable.   The error message is "super_num devices 1 mismatch with num_devices 1 found here" on the primary cache drive.   If you look at /var/log/syslog during the array mount process you get this:

Apr  9 23:04:44 Tower emhttpd: shcmd (100): mkdir -p /mnt/cache
Apr  9 23:04:44 Tower emhttpd: /mnt/cache uuid: 54f4fb0c-8643-4e1b-83de-a1566868fc69
Apr  9 23:04:44 Tower emhttpd: /mnt/cache TotDevices: 2
Apr  9 23:04:44 Tower emhttpd: /mnt/cache NumDevices: 2
Apr  9 23:04:44 Tower emhttpd: /mnt/cache NumFound: 2
Apr  9 23:04:44 Tower emhttpd: /mnt/cache NumMissing: 0
Apr  9 23:04:44 Tower emhttpd: /mnt/cache NumMisplaced: 0
Apr  9 23:04:44 Tower emhttpd: /mnt/cache NumExtra: 0
Apr  9 23:04:44 Tower emhttpd: /mnt/cache LuksState: 0
Apr  9 23:04:44 Tower emhttpd: shcmd (101): mount -t btrfs -o noatime,space_cache=v2,discard=async -U 54f4fb0c-8643-4e1b-83de-a1566868fc69 /mnt/cache
Apr  9 23:04:44 Tower kernel: BTRFS info (device sdf1): turning on async discard
Apr  9 23:04:44 Tower kernel: BTRFS info (device sdf1): using free space tree
Apr  9 23:04:44 Tower kernel: BTRFS info (device sdf1): has skinny extents
Apr  9 23:04:44 Tower root: mount: /mnt/cache: wrong fs type, bad option, bad superblock on /dev/sdg1, missing codepage or helper program, or other error.
Apr  9 23:04:44 Tower kernel: BTRFS error (device sdf1): super_num_devices 1 mismatch with num_devices 1 found here
Apr  9 23:04:44 Tower kernel: BTRFS error (device sdf1): failed to read chunk tree: -22
Apr  9 23:04:44 Tower kernel: BTRFS error (device sdf1): open_ctree failed
Apr  9 23:04:44 Tower emhttpd: shcmd (101): exit status: 32
Apr  9 23:04:44 Tower emhttpd: /mnt/cache mount error: No file system
Apr  9 23:04:44 Tower emhttpd: shcmd (102): umount /mnt/cache
Apr  9 23:04:44 Tower root: umount: /mnt/cache: not mounted.
Apr  9 23:04:44 Tower emhttpd: shcmd (102): exit status: 32
Apr  9 23:04:44 Tower emhttpd: shcmd (103): rmdir /mnt/cache

 

I have 2 cache drives - they both show up as un-mountable when this problem appears.    One thing I tried tonight was disconnecting one of the cache drives (primary) - booted up - same issue but not seeing the primary drive - which is expected.   Shutdown - re-connected drive - booted up - same error but now with a warning that the primary cache drive will be over written.  Rebooted - the warning went away but back to the un-mounted problem.  Shut down - rolled the SATA cables on the cache drives - booted back up - started array - now the cache drives both mount.   Just a note.  I started to have this problem AFTER I moved my drives to a newer server.   Right after I moved my drives - and saw that un-mountable error - I shut down - moved everything back to the old server - powered up - started array - cache drives mount normally.  Shutdown - move to new system - boot up - cache drives un-mountable.   So far I've been able to get the cache drives to mount by trying various things - but there is no consistency to what works.  Tonight when I rolled the SATA cables it came up - but I don't know if that did it or the stars aligned under a full moon.  Anyway if anyone has any suggestions...

 

...REK

   

 

 

Very strange issue, nest time it doesn't mount please post the complete diagnostics.

Will do.   Thanks for the response...