So a disk from a cache pool has failed. 
The pool is (RAID 1 I believe?), consisting out of 2 number of 1TB nvme ssd and 2 numbers of 2TB nvme ssd. 

 

Having rebooted unraid, unraid now says

BTRFS info (device dm-3): relocating block group 1583833153536 flags data|raid1

and subsequently 

btrfs-no-space-left-errno-28

and 

BTRFS info (device dm-3: state EA): forced readonly

I have backups using kopia, which I can restore.

 

Was trying to move the files from /mnt/cache to /mnt/disk2, but having issues with io errors etc. 

 

Thinking I now have 2 options:
1. borrow a 1TB SSD and add it to the pool, and let the btrfs sort itself out. 

2. delete everything in cache pool and recover from backup. 

 

PS: 

Made a mistake and tried to fix things, and rebooted unraid before grabbing the diagnostics. 

 

PPS:

Was reading on the forum and wondering if the commands mentioned will help, but figured would ask here first. 

 

 

 

athena-diagnostics-20240831-0959.zip

You can try to manually mount the pool with the skip_balance option, then move some data to free up some space, then mount normally to resume the balance, let me know if you need help with the commands.

Yes, please. I would need help with the commands. 

Sorry, some other questions. 
1. In my case where the data is approximately 2tb, how much more space would the balance require? 
2. Also, I was lucky that this happened the last few hours before the weekend, but I would need to make the server available to the company by Monday. While unraid performs the balance, can the server be accessed as normal? 

.... and I'm not sure what happened to cause this. 🥲

Didn't notice before that the pool is encrypted, that complicates things by a lot, and since I'm about to go out for the weekend, won't have the opportunity to follow up, since you have a backup I would recommend re-formatting and restoring from there.

Gotcha. Thanks!