I create a new BTrFS raid10 file system using two 250GB drives and the second partition on a third 80GB drive. I create a subvol and snapshot. I mount the snapshot and start copying 8GB of data to it. It gets to around 1GB and the Desktop disappears and what looks like a non interactive terminal comes up with dump/crash information. I don't have a camera handy or I'd take a picture and post it. It basically looks like stack trace info. CTRL-ALT F7 will eventually bring back the Desktop though but the entire BTrFS portion of the OS is hung and non responsive until I reboot.
I've reformated and reproduced this problem 3 times now and I'm about to give up :(
I realize it is possible this problem is not entirely BTrFS' fault because I'm on natty which is still alpha.
More granular details in case I'm an idiot:
1) Create FS: sudo mkfs.btrfs -m raid10 -d raid10 /dev/sda2 /dev/sdb /dev/sdc 2) Initial temporary mount: mkdir /btrfs && sudo mount -t btrfs /dev/sda2 /btrfs 3) Create subvol btrfs s c /btrfs/vm 4) Create initial snapshot: (optional) btrfs s sn /btrfs/cantremember.snap.something 5)unmount /btrfs and mount /btrfs/vm sudo mount -t btrfs -o subvol=vm /dev/sda2 /btrfs/vm 6) Copy data to subvolume. 7) Balance data across drives: (optional) btrfs f bal <path>
(never get to this step 7...) Am I doing something wrong?
EDIT: I managed to catch the tail end of the backtrace / crash info:
kernel BUG at /build/buildd/linux-2.6.38/fs/btrfs/extent-tree.c:8581
EDIT2: Removing the smallest (46GB) partition from the raid10 array seems to have eliminated the problem.