On 02/09/13 15:06, Stefan Sperling wrote:
On Sat, Feb 09, 2013 at 03:52:12AM -0500, Scott McEachern wrote:
On 02/09/13 03:09, Andy Bradford wrote:
Thus said Joel Sing on Sat, 09 Feb 2013 16:44:11 +1100:
umount via DUID does not work currently - this will be fixed shortly
after the next release freeze has ended.
Will that also include shutdown of softraid via DUID? e.g.,
bioctl -d DUID
Or is this not even possible?
Thanks,
Andy
Oddly enough, no.
See http://marc.info/?l=openbsd-tech&m=133513662106783&w=2 for a patch.
It hasn't been committed yet because jsing didn't ok it. Perhaps he
will change his mind if we ask again nicely :)
The patch applied cleanly, I rebuilt the system and rebooted. All
looked good.
Then I adjusted my /etc/rc.shutdown to this:
umount -f /st7
umount -f /home
#bioctl -d sd10 <-- this was used before
bioctl -d 485a9f963f9cf9ea
#bioctl -d 485a9f963f9cf9ea.a
#bioctl -d sd11 <-- this was used before
bioctl -d 36d18f2cde909b01
#bioctl -d 36d18f2cde909b01.a
and executed a reboot.
The bad news? I got the same error as before:
syncing disks... done
sd3 detached
softraid0: I/O error 5 on dev 0x433 at block 16
softraid0: could not write metadata to sd3d
sd4 detached
rebooting...
at least I think that's what it said, it went by rather quickly. I
definitely saw the "could not write metadata" part.
At this point I figured no harm, no foul. Was I ever wrong.
Upon reboot the system shit all over the place and dropped me to single
user mode. The offending partitions were /dev/sd8a and /dev/sd9a. In
my fstab, I have the following:
6be798121798a5a7.b none swap sw
6be798121798a5a7.a / ffs rw,softdep 1 1
6be798121798a5a7.d /tmp ffs rw,nodev,nosuid,softdep 1 2
6be798121798a5a7.f /usr ffs rw,nodev,softdep 1 2
6be798121798a5a7.g /usr/X11R6 ffs rw,nodev,softdep 1 2
6be798121798a5a7.i /usr/local ffs rw,nodev,softdep 1 2
6be798121798a5a7.h /usr/obj ffs rw,nodev,nosuid,softdep 1 2
6be798121798a5a7.e /var ffs rw,nodev,nosuid,softdep 1 2
e1d635ac777ed919.a /st5 ffs rw,nodev,nosuid,noexec,noatime,softdep 1 2
3131dc858bdefd32.a /st6 ffs rw,nodev,nosuid,noexec,noatime,softdep 1 2
darkon:/st1/ /st1 nfs rw,nodev,soft,intr 0 0
See the /st5 (e1d..919.a, aka sd8a) and /st6 (313..f32.a, aka sd9a)
mount points? Those are my two 3TB RAID1 volumes. Or should I say,
*were*. You can see where this is going, right?
I used ed(1) to comment those lines out, rebooted. Things seemed to
come up normally and I figured I might have to fsck the big drives
when.... oh *fuck*. sd8 and sd9 no longer exist.
The tail end of my dmesg normally looks like this (before I added the
crypto volumes):
softraid0 at root
scsibus4 at softraid0: 256 targets
sd8 at scsibus4 targ 1 lun 0: <OPENBSD, SR RAID 1, 005> SCSI2 0/direct fixed
sd8: 2861588MB, 512 bytes/sector, 5860532576 sectors
sd9 at scsibus4 targ 2 lun 0: <OPENBSD, SR RAID 1, 005> SCSI2 0/direct fixed
sd9: 2861588MB, 512 bytes/sector, 5860532576 sectors
root on sd5a (6be798121798a5a7.a) swap on sd5b dump on sd5b
Now it looks like this:
softraid0 at root
scsibus4 at softraid0: 256 targets
root on sd5a (6be798121798a5a7.a) swap on sd5b dump on sd5b
I didn't know what to wipe first, the sweat off my forehead or ... well,
you get the idea.
I'm tempted to try to use "bioctl -c 1 -l /dev/sd0,/dev/sd1 softraid0"
and "bioctl -c 1 -l /dev/sd2,/dev/sd3 softraid0" to recreate the volumes
(just like how I created them the first time around), and *hope like
hell* I can get my shit back, but before I do that, I wanted to get your
advice to ensure that's my best possible move.
Hey, you know, maybe it would be best if I reinstalled my previous
snapshot (Feb7 I think) and use _that_ version of bioctl, no?
--
Scott McEachern
https://www.blackstaff.ca