zfs: panic on import
jtotz at imperial.ac.uk
Tue Nov 15 19:31:41 UTC 2011
On 15/11/2011 19:24, Freddie Cash wrote:
> On Tue, Nov 15, 2011 at 11:14 AM, Johannes Totz<jtotz at imperial.ac.uk>wrote:
>> I get a lovely panic everytime i try to import my pool.
>> panic: page fault
>> #5 ... calltrap
>> #6 ... zio_vdev_io_start
>> #7 ... zio_execute
>> #8 ... zio_ioctl
>> #9 ... zio_flush
>> #10 ... vdev_config_sync
>> #11 ... spa_sync
>> #12 ... txg_sync_thread
>> #13 ... fork_exit
>> #14 ... fork_trampoline
>> This is for a funky config in which one half of a zfs-mirror sits on a
>> local gpt partition and the other half on a local file on a ufs volume
>> (dont ask why).
>> The machine is running a 8.2-stable, compiled on 5th Sept 2011.
>> Maybe r226617 fixed this?
>> Any idea how to get it back alive?
> Have you tried "import -F" or to import it readonly?
Also, I can delete/rename the second-half-on-ufs of the mirror but that
gets me the same panic; instead of just a degraded pool.
More information about the freebsd-fs