Disaster? After detaching my pool, I get shared object errors and cannot boot

[sunshine931]

I'm in the process of redoing my pool layout. My pool "nebula" was using disks I want for the new layout, and I needed to free those disks up.

I created a new pool called "md1000", did zfs send/receive to get my data there, then scrubbed it and reviewed SMART data.

After I was certain my data was safe, I detached "nebula" via the UI, preserving share config, and destroying data. I planned to create a new pool with the name nebula and figured this could save me from having to redo my iSCSI and NFS shares.

Immediately I realized something has gone wrong - shared object errors on virtually every command. I couldn't even shutdown or reboot, so I powered off thinking a reboot might correct whatever went awry. It has not helped.

What the heck happened?! How do I recover? Thanks!

 

[SweetAndLow]

Reinstall, create new pool with correct name, copy data over and finally upload your config file.

 

[Apollo]

@sunshine931, Instead of reinstalling, if md1000 is attachable, you could try the export command and import it with the new name. Make sure you are not using encrypted pools.

https://forums.freenas.org/index.php?threads/renaming-volume-with-data-already-in-it.13061/

 

[sunshine931]

Thanks everyone. I'm taking the "reinstall and zfs send/ receive to new pool" approach.

I'm curious though - what happened?! I'd expect that detaching my pool should have no impact on FreeNAS's operation. Why would that blow the OS up?

 

[SweetAndLow]

It has something to do with how you didn't delete the share config I would assume.

 

[sunshine931]

It has something to do with how you didn't delete the share config I would assume.

I'm thinking this might be worthy of a bug/ RFI. It's certainly not the expected behaviour.