On 1 Jul 2010, at 05:51, Daniel Kozlowski wrote: > I am also having the same problem with a slightly different setup. In My case I > cannot mount the filesystem. mount, btrfs-endio-met and kblockd/0 will all > continually run until the system freezes up and requires a power cycle. have you tried mounting with '-o degraded'? having monitored the system for a while, i also think that in fact it's btrfs that's killing my system. i'm on ubuntu 10.4 with: $ uname -a Linux htpc 2.6.32-22-server #36-Ubuntu SMP Thu Jun 3 20:38:33 UTC 2010 x86_64 GNU/Linux using the default kernel module, but git'd out the tools. following the other thread 'Is there a more aggressive fixer than btrfsck?' i suspect that we'll just have to wait until some actual fsck operations are available for btrfs :( on my system, it's btrfs-endio-met (only 1 out of 4) and btrfs-transacti (1 out of 2) that is taking up all the cpu/io wait cycles. i wonder if it's only certain files on the array that are hosed; if that's the case is there a way i can map the kernel messages to a real filename? i don't mind loosing the odd file on this array, but i don't fancy copying it all over to somewhere else (yeah-yeah, up to date backups blah blah!) - i figured given the momentum btrfs was gaining it would be much more stable than this :( Yee.-- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
