Henrik Johansson
2009-Sep-13 21:29 UTC
[zfs-discuss] Error recovery with replicated metadata
Hello all,
I have managed to get my hands on a OSOL 2009.06 root disk which has
three failed blocks on it, these three blocks makes it impossible to
boot from the disk and to import the disk to on another machine. I
have checked the disk and three blocks are inaccessible, quite close
to each other. Now should this not have good a good chance of being
saved by replicated metadata? The data on the disk is usable, i did a
block copy of the whole disk to a new one, and the scrub works out
flawlessly. I guess this could be a timeout issue, but the disk is at
least a WD RE2 disk with error recovery of 7 seconds. The failing
systems release was 111a, and I have tried to import it into 122.
The disk was used by one of my friends which i have converted into
using Solaris and ZFS for his company storage needs, and he is a bit
skeptical when three blocks makes the whole pool unusable. The good
part is that he uses mirrors for his rpool even on this non critical
system now ;)
Anyway, can someone help to explain this, is there any timeouts that
can be tuned to import the pool or is this a feature, obviously all
data that is needed is intact on the disk since the block copy of the
pool worked fine.
Also don''t we need a force option for the -e option to zdb, so that we
can use it with pools thats not have been exported correctly from a
failing machine?
The import timeouts after 41 seconds:
root at arne:/usr/sbin# zpool import -f 2934589927925685355 dpool
cannot import ''rpool'' as ''dpool'': one or
more devices is currently
unavailable
root at arne:/usr/sbin# zpool import
pool: rpool
id: 2934589927925685355
state: ONLINE
status: The pool was last accessed by another system.
action: The pool can be imported using its name or numeric identifier
and
the ''-f'' flag.
see: http://www.sun.com/msg/ZFS-8000-EY
config:
rpool ONLINE
c1t4d0s0 ONLINE
Damaged blocks as reported by format:
Medium error during read: block 8646022 (0x83ed86) (538/48/28)
ASC: 0x11 ASCQ: 0x0
Medium error during read: block 8650804 (0x840034) (538/124/22)
ASC: 0x11 ASCQ: 0x0
Medium error during read: block 8651987 (0x8404d3) (538/143/8)
ASC: 0x11 ASCQ: 0x0
What i managed to get out of zdb:
root at arne:/usr/sbin# zdb -e 2934589927925685355
WARNING: pool ''2934589927925685355'' could not be loaded as it
was last
accessed by another system (host: keeper hostid: 0xc34967). See:
http://www.sun.com/msg/ZFS-8000-EY
zdb: can''t open 2934589927925685355: No such file or directory
root at arne:/usr/sbin# zdb -l /dev/dsk/c1t4d0s0
--------------------------------------------
LABEL 0
--------------------------------------------
version=14
name=''rpool''
state=0
txg=269696
pool_guid=2934589927925685355
hostid=12798311
hostname=''keeper''
top_guid=9161928630964440615
guid=9161928630964440615
vdev_tree
type=''disk''
id=0
guid=9161928630964440615
path=''/dev/dsk/c7t1d0s0''
devid=''id1,sd at
SATA_____WDC_WD5000YS-01M_____WD-WCANU2080316/a''
phys_path=''/pci at 0,0/pci8086,2847 at 1c,4/pci1043,81e4 at 0/
disk at 1,0:a''
whole_disk=0
metaslab_array=23
metaslab_shift=32
ashift=9
asize=500067467264
is_log=0
--------------------------------------------
LABEL 1
--------------------------------------------
version=14
name=''rpool''
state=0
txg=269696
pool_guid=2934589927925685355
hostid=12798311
hostname=''keeper''
top_guid=9161928630964440615
guid=9161928630964440615
vdev_tree
type=''disk''
id=0
guid=9161928630964440615
path=''/dev/dsk/c7t1d0s0''
devid=''id1,sd at
SATA_____WDC_WD5000YS-01M_____WD-WCANU2080316/a''
phys_path=''/pci at 0,0/pci8086,2847 at 1c,4/pci1043,81e4 at 0/
disk at 1,0:a''
whole_disk=0
metaslab_array=23
metaslab_shift=32
ashift=9
asize=500067467264
is_log=0
--------------------------------------------
LABEL 2
--------------------------------------------
version=14
name=''rpool''
state=0
txg=269696
pool_guid=2934589927925685355
hostid=12798311
hostname=''keeper''
top_guid=9161928630964440615
guid=9161928630964440615
vdev_tree
type=''disk''
id=0
guid=9161928630964440615
path=''/dev/dsk/c7t1d0s0''
devid=''id1,sd at
SATA_____WDC_WD5000YS-01M_____WD-WCANU2080316/a''
phys_path=''/pci at 0,0/pci8086,2847 at 1c,4/pci1043,81e4 at 0/
disk at 1,0:a''
whole_disk=0
metaslab_array=23
metaslab_shift=32
ashift=9
asize=500067467264
is_log=0
--------------------------------------------
LABEL 3
--------------------------------------------
version=14
name=''rpool''
state=0
txg=269696
pool_guid=2934589927925685355
hostid=12798311
hostname=''keeper''
top_guid=9161928630964440615
guid=9161928630964440615
vdev_tree
type=''disk''
id=0
guid=9161928630964440615
path=''/dev/dsk/c7t1d0s0''
devid=''id1,sd at
SATA_____WDC_WD5000YS-01M_____WD-WCANU2080316/a''
phys_path=''/pci at 0,0/pci8086,2847 at 1c,4/pci1043,81e4 at 0/
disk at 1,0:a''
whole_disk=0
metaslab_array=23
metaslab_shift=32
ashift=9
asize=500067467264
is_log=0
Regards
Henrik
http://sparcv9.blogspot.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20090913/45ff365e/attachment.html>