My current interpretation of this problem is that it is some
pathological condition caused by not rebalancing and being nearly out
of space for allocating more metadata and hence it is rarely being
seen by anyone else (because most users are regularly doing
rebalances).
See this thread for details
I've reproduced these issues on a single-core machine which doesn't
appear to become completely unresponsive after 12 hours of copying (as
the other machines are deadlocking after 5-10 minutes, perhaps?), but
it does use 100% SYS CPU with no IO traffic for the vast majority of
the time. (In fact, c
I should add that I have reproduced this even after doing `mount -o
clear_cache /dev/... /mnt/...`, unmount, remount with `-o
space_cache`. After the machine lockup and rebooting there are the
warnings of the form:
> [ 117.288248] BTRFS warning (device dm-0): block group 694165700608 has wrong
>
I've now reproduced this on 3.15.7-031507-generic and
3.16.0-031600rc7-generic, and have a test case where I can reliably
cause the crash after about 30 seconds of disk activity.
The test case just involves taking a directory tree of ~400GB of files
and copying every file to a new one with .new on
The crashes became more frequent. The time scale before lockup went
~12 days, ~7 days, ~2 days, ~6 hours, ~1 hour.
Then we upgraded to 3.15.7-031507-generic on the advice of
#ubuntu-kernel and #btrfs on IRC, and it has since been stable for
19.5 hours.
I dug around more in our logs and realised t
Someone on IRC suggested that I clear the free cache:
> sudo mount -o remount,clear_cache /path/to/dev /path/to/mount
> sudo mount -o remount,space_cache /path/to/dev /path/to/mount
The former command printed `btrfs: disk space caching is enabled` and
the latter repeated it, making me think that
Hi All,
I've reported a bug with Ubuntu here:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1349711
The machine in question has one BTRFS volume which is 87% full and
lives on an Logical Volume Manager (LVM) block device on top of one
Amazon Elastic Block Store (EBS) device.
We have other
