Error after removing cache drives: "Unmountable: No pool uuid"

[Phillycj]

Hi, I updated to 6.9 stable today and I tried taking advantage of multiple pools and tried removing 2 of my 4 SSDs from the cache pool. I spun down the array, reset the config for the cache pool and removed the bottom two drives, but did not bring the number of the cache pool down. Upon booting back up, I am constantly getting the "Unmountable: No pool uuid" error across all of my cache drives, even after readding the two removed drives. I tried looking up the error but the one forum post I could find with this error wasn't all that helpful for my scenario. Is anyone able to give me a solution? Thanks!

sol-diagnostics-20210302-1848.zip

[JorgeB]

18 minutes ago, Phillycj said:

reset the config for the cache pool and removed the bottom two drives

Not quite clear what you mean by this, did the pool have data? You can't remove two devices at once and maintain data.

[Phillycj]

1 minute ago, JorgeB said:

Not quite clear what you mean by this, did the pool have data? You can't remove two devices at once and maintain data.

The pool has data, yes. I have not attempted to format any of the drives yet.

[JorgeB]

3 minutes ago, Phillycj said:

I have not attempted to format any of the drives yet.

 

But you unassigned two cache devices and started the array, so they were both wiped:

 

Mar  2 18:40:43 Sol emhttpd: shcmd (1232): /sbin/wipefs -a /dev/sde1
...
Mar  2 18:40:43 Sol emhttpd: shcmd (1234): /sbin/wipefs -a /dev/sdf1

 

And like mentioned you can' remove two devices at the same time and keep the pool, it needs to be one device at a time (assuming a raid1 or raid10 pool).

[Phillycj]

6 minutes ago, JorgeB said:

 

But you unassigned two cache devices and started the array, so they were both wiped:

 

Mar  2 18:40:43 Sol emhttpd: shcmd (1232): /sbin/wipefs -a /dev/sde1
...
Mar  2 18:40:43 Sol emhttpd: shcmd (1234): /sbin/wipefs -a /dev/sdf1

 

And like mentioned you can' remove two devices at the same time and keep the pool, it needs to be one device at a time (assuming a raid1 or raid10 pool).

If just the file system on the two drives was wiped, is there any way to restore the fs on these drives while keeping the contents intact?

[JorgeB]

Very possibly, but beyond my knowledge, you can likley find some help for that using #btrfs on freenode IRC.

[Phillycj]

Just now, JorgeB said:

Very possibly, but beyond my knowledge, you can likley find some help for that using #btrfs on freenode IRC.

Great, thanks for the help!