I recently created a RAID-Z1 pool out of a set of 7 SCSI disks, using the following command:
# zpool create magicant raidz c5t0d0 c5t1d0 c5t2d0 c5t3d0 c5t4d0 c5t5d0 c5t6d0
It worked fine, but I was slightly confused by the size yield (99 GB vs the 116 GB I had on my other RAID-Z1 pool of same-sized disks).
I thought one of the disks might have been to blame, so I tried swapping it out - it turned out my replacement disk was a dud (zpool wasn't happy about that, and eventually offlined the disk). Oh well, swap the old one back in, no harm done.
Reboot, and ZFS informs me that I'm missing another, unrelated disk (c5t1d0 was the one I tried swapping out unsuccessfully - c5t3d0 is the one it complained about, which I had swapped for another disk before any problems began or any data was in the pool, with no problems - ZFS scrubbed and was happy).
It continually claimed the device was unavailable and so the pool was in degraded mode - attempting to replace the disk with itself yielded an error about the disk being in use by the same pool which claimed the disk was unavailable. Unmount the pool, same error persists, zpool replace continues to give that error, despite repeated zpool offline magicant c5t3d0 followed by zpool online [etc].
I try exporting and re-importing the pool - the export went fine. The import threw the confusing error which is the point of this email:
# zpool import
pool: magicant
id: 3232403590553596936
state: FAULTED
status: One or more devices are missing from the system.
action: The pool cannot be imported. Attach the missing
devices and try again.
see: http://www.sun.com/msg/ZFS-8000-6X
config:
magicant UNAVAIL missing device
raidz1 ONLINE
c5t0d0 ONLINE
c5t1d0 ONLINE
c5t2d0 ONLINE
c5t3d0 ONLINE
c5t4d0 ONLINE
c5t5d0 ONLINE
c5t6d0 ONLINE
Additional devices are known to be part of this pool, though their
exact configuration cannot be determined.
So, to summarize:
7-scsi-disk raidz1 zpool is created
c5t3d0 is swapped out for another disk of identical size, zfs is happy and functions fine after a scrub
c5t1d0 is swapped out for another disk of identical size (which happened to be a dud), Solaris didn't like that, so I put the original back in and rebooted
On boot, zpool claims c5t3d0 is unavailable, while format and cfgadm both agree that the disk still exists and is dandy. zpool replace pool c5t3d0 c5t3d0 claims it's in use by that pool, zpool offline pool c5t3d0 followed by zpool online pool c5t3d0 doesn't help. zpool export pool worked, but then zpool import pool threw the above error.
Is this a bug, or am I missing something obvious?
snv 44, x86.
- Rich
_______________________________________________ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
