I've decided to upgrade my home server capacity by replacing the disks in one
of my mirror vdevs. The procedure appeared to work out, but during resilver, a
couple million checksum errors were logged on the new device. I've read through
quite a bit of the archive and searched around a bit, but can not find anything
definitive to ease my mind on whether to proceed.
SunOS deepthought 5.10 Generic_142901-13 i86pc i386 i86pc
pool: tank
state: DEGRADED
status: One or more devices is currently being resilvered. The pool will
continue to function, possibly in a degraded state.
action: Wait for the resilver to complete.
scrub: resilver in progress for 0h0m, 0.00% done, 691h28m to go
config:
NAME STATE READ WRITE CKSUM
tank DEGRADED 0 0 0
mirror DEGRADED 0 0 0
replacing DEGRADED 215 0 0
c1t6d0s0/o FAULTED 0 0 0 corrupted data
c1t6d0 ONLINE 0 0 215 3.73M resilvered
c1t2d0 ONLINE 0 0 0
mirror ONLINE 0 0 0
c1t1d0 ONLINE 0 0 0
c1t5d0 ONLINE 0 0 0
mirror ONLINE 0 0 0
c1t0d0 ONLINE 0 0 0
c1t4d0 ONLINE 0 0 0
logs
c8t1d0p1 ONLINE 0 0 0
cache
c2t1d0p2 ONLINE 0 0 0
During the resilver, the cache device and the zil were both removed for errors
(1-2k each). (Despite the c2/c8 discrepancy, they are partitions on the same
OCZvertexII device.)
# zpool status -xv tank
pool: tank
state: DEGRADED
status: One or more devices has experienced an unrecoverable error. An
attempt was made to correct the error. Applications are unaffected.
action: Determine if the device needs to be replaced, and clear the errors
using 'zpool clear' or replace the device with 'zpool replace'.
see: http://www.sun.com/msg/ZFS-8000-9P
scrub: resilver completed after 9h20m with 0 errors on Sat Jun 19 22:07:27 2010
config:
NAME STATE READ WRITE CKSUM
tank DEGRADED 0 0 0
mirror ONLINE 0 0 0
c1t6d0 ONLINE 0 0 2.69M 539G resilvered
c1t2d0 ONLINE 0 0 0
mirror ONLINE 0 0 0
c1t1d0 ONLINE 0 0 0
c1t5d0 ONLINE 0 0 0
mirror ONLINE 0 0 0
c1t0d0 ONLINE 0 0 0
c1t4d0 ONLINE 0 0 0
logs
c8t1d0p1 REMOVED 0 0 0
cache
c2t1d0p2 REMOVED 0 0 0
I cleared the errors (about 5000/GB resilvered!), removed the cache device, and
replaced the zil partition with the whole device. After 3 pool scrubs with no
errors, I want to check with someone else that it appears okay to replace the
second drive in this mirror vdev. The one thing I have not tried is a large
file transfer to the server, as I am also dealing with an NFS mount problem
which popped up suspiciously close to my most recent patch update.
# zpool status -v tank
pool: tank
state: ONLINE
scrub: scrub completed after 3h26m with 0 errors on Mon Jun 21 01:45:00 2010
config:
NAME STATE READ WRITE CKSUM
tank ONLINE 0 0 0
mirror ONLINE 0 0 0
c1t6d0 ONLINE 0 0 0
c1t2d0 ONLINE 0 0 0
mirror ONLINE 0 0 0
c1t1d0 ONLINE 0 0 0
c1t5d0 ONLINE 0 0 0
mirror ONLINE 0 0 0
c1t0d0 ONLINE 0 0 0
c1t4d0 ONLINE 0 0 0
logs
c0t0d0 ONLINE 0 0 0
errors: No known data errors
/var/adm/messages is positively over-run with these triplets/quadruplets, not
all of which end which end up as "fatal" type.
Jun 19 21:43:19 deepthought scsi: [ID 107833 kern.warning] WARNING:
/p...@0,0/pci1043,8...@7/d...@1,0 (sd14):
Jun 19 21:43:19 deepthought Error for Command: write(10)
Error Level: Retryable
Jun 19 21:43:19 deepthought scsi: [ID 107833 kern.notice] Requested
Block: 26721062 Error Block: 26721062
Jun 19 21:43:19 deepthought scsi: [ID 107833 kern.notice] Vendor: ATA
Serial Number:
Jun 19 21:43:19 deepthought scsi: [ID 107833 kern.notice] Sense Key:
Aborted Command
Jun 19 21:43:19 deepthought scsi: [ID 107833 kern.notice] ASC: 0x8 (LUN
communication failure), ASCQ: 0x0, FRU: 0x0
Jun 19 21:43:19 deepthought scsi: [ID 107833 kern.warning] WARNING:
/p...@0,0/pci1043,8...@7/d...@1,0 (sd14):
Jun 19 21:43:19 deepthought Error for Command: write(10)
Error Level: Retryable
Jun 19 21:43:19 deepthought scsi: [ID 107833 kern.notice] Requested
Block: 26721062 Error Block: 26721062
Jun 19 21:43:19 deepthought scsi: [ID 107833 kern.notice] Vendor: ATA
Serial Number:
Jun 19 21:43:19 deepthought scsi: [ID 107833 kern.notice] Sense Key:
Aborted Command
Jun 19 21:43:19 deepthought scsi: [ID 107833 kern.warning] WARNING:
/p...@0,0/pci1043,8...@7/d...@1,0 (sd14):
Jun 19 21:43:19 deepthought Error for Command: write(10)
Error Level: Fatal
Jun 19 21:43:19 deepthought scsi: [ID 107833 kern.notice] Requested
Block: 26721062 Error Block: 26721062
Jun 19 21:43:19 deepthought scsi: [ID 107833 kern.notice] Vendor: ATA
Serial Number:
Jun 19 21:43:19 deepthought scsi: [ID 107833 kern.notice] Sense Key:
Aborted Command
In the past this kernel.notice ID has come up as "informational" for others,
and for my case it _only_ occurred during the initial resilver. One last point
of interest is the new drive is the WD Green WD10EARS, and the old are WD Green
WD6400AACS (all of which I have tested on another system with the WD read-test
utility). I know these drives get their share of ridicule (and occasional
praise/satisfaction), but I'd appreciate any thoughts on proceeding with the
mirror upgrade. [Backups are a check.]
Justin
_______________________________________________
zfs-discuss mailing list
zfs-discuss@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
