ZFS mount freezing

S

spamx

Hello everybody!

My configuration:
3x400GB Single-parity RAID-5 (raidz1)

bsd# zpool list
Code: 
NAME        SIZE   USED  AVAIL    CAP  HEALTH  ALTROOT
ZFS-Vol01  1.16T   596G   596G    50%  ONLINE  -
bsd# zpool status -v
Code: 
  pool: ZFS-Vol01
 state: ONLINE
 scrub: none requested
config:

        NAME        STATE     READ WRITE CKSUM
        ZFS-Vol01   ONLINE       0     0     0
          raidz1    ONLINE       0     0     0
            da3     ONLINE       0     0     0
            da2     ONLINE       0     0     0
            da1     ONLINE       0     0     0

errors: No known data errors
bsd# zfs list
Code: 
NAME        USED  AVAIL  REFER  MOUNTPOINT
ZFS-Vol01   397G   384G   397G  /mnt/l
bsd# zfs list
Code: 
NAME        USED  AVAIL  REFER  MOUNTPOINT
ZFS-Vol01   397G   384G   397G  /mnt/l
bsd# ls -as /mnt/l
Code: 
total 4
2 .     2 ..
BLANK DIR !

bsd# mount
Code: 
/dev/da0s1a on / (ufs, local)
devfs on /dev (devfs, local, multilabel)
/dev/da0s1e on /tmp (ufs, local, soft-updates)
/dev/da0s1f on /usr (ufs, local, soft-updates)
/dev/da0s1d on /var (ufs, local, soft-updates)

bsd# zfs mount -a
Code: 
load: 0.00  cmd: zfs 76483 [tx->tx_quiesce_done_cv)] 8.11r 0.00u 0.08s 0% 1704k
load: 0.00  cmd: zfs 76483 [tx->tx_quiesce_done_cv)] 9.92r 0.00u 0.08s 0% 1704k
load: 0.00  cmd: zfs 76483 [tx->tx_quiesce_done_cv)] 10.45r 0.00u 0.08s 0% 1704k
load: 0.00  cmd: zfs 76483 [tx->tx_quiesce_done_cv)] 10.86r 0.00u 0.08s 0% 1704k
load: 0.00  cmd: zfs 76483 [tx->tx_quiesce_done_cv)] 14.73r 0.00u 0.08s 0% 1704k
load: 0.00  cmd: zfs 76483 [tx->tx_quiesce_done_cv)] 15.05r 0.00u 0.08s 0% 1704k
load: 0.00  cmd: zfs 76483 [tx->tx_quiesce_done_cv)] 472.47r 0.00u 0.08s 0% 1704k
load: 0.00  cmd: zfs 76483 [tx->tx_quiesce_done_cv)] 472.77r 0.00u 0.08s 0% 1704k
load: 0.00  cmd: zfs 76483 [tx->tx_quiesce_done_cv)] 473.26r 0.00u 0.08s 0% 1704k
load: 0.00  cmd: zfs 76483 [tx->tx_quiesce_done_cv)] 606.34r 0.00u 0.08s 0% 1704k
load: 0.00  cmd: zfs 76483 [tx->tx_quiesce_done_cv)] 606.62r 0.00u 0.08s 0% 1704k
load: 0.00  cmd: zfs 76483 [tx->tx_quiesce_done_cv)] 606.93r 0.00u 0.08s 0% 1704k
load: 0.00  cmd: zfs 76483 [tx->tx_quiesce_done_cv)] 607.21r 0.00u 0.08s 0% 1704k
load: 0.00  cmd: zfs 76483 [tx->tx_quiesce_done_cv)] 656.75r 0.00u 0.08s 0% 1704k
load: 0.00  cmd: zfs 76483 [tx->tx_quiesce_done_cv)] 657.12r 0.00u 0.08s 0% 1704k
load: 0.00  cmd: zfs 76483 [tx->tx_quiesce_done_cv)] 657.52r 0.00u 0.08s 0% 1704k
CTRL+T output

I try mount pool (zfs mount -a), then zfs command is freezing! I can only reboot with ctrl+alt+del

Please help

P.S. I have no backups... :-(

Thanks a lot!
 
I´m not really sure what your errors mean, but I guess you´re hoping someone will suggest something! So here you go:

Is it really still showing "no known errors" after running the failed mount command? Have you checked the drive status from "camcontrol devlist"? And maybe it would be worth doing an export and import of the pool to see if that cleans up the situation?

ta Andy.
 
Thx for your quick reply.

The [cmd=]camcontrol devlist[/cmd] shows all drive

Code: 
bsd# zpool export ZFS-Vol01
command successfully go all pool exported

Code: 
bsd# zpool import ZFS-Vol01
load: 0.01  cmd: zpool 1051 [tx->tx_quiesce_done_cv)] 242.65r 0.00u 0.13s 0% 2176k
load: 0.01  cmd: zpool 1051 [tx->tx_quiesce_done_cv)] 246.21r 0.00u 0.13s 0% 2176k
load: 0.01  cmd: zpool 1051 [tx->tx_quiesce_done_cv)] 246.51r 0.00u 0.13s 0% 2176k
load: 0.01  cmd: zpool 1051 [tx->tx_quiesce_done_cv)] 246.83r 0.00u 0.13s 0% 2176k
load: 0.01  cmd: zpool 1051 [tx->tx_quiesce_done_cv)] 247.32r 0.00u 0.13s 0% 2176k
load: 0.01  cmd: zpool 1051 [tx->tx_quiesce_done_cv)] 247.70r 0.00u 0.13s 0% 2176k
load: 0.01  cmd: zpool 1051 [tx->tx_quiesce_done_cv)] 248.19r 0.00u 0.13s 0% 2176k

then zpool import command is freezing! I can only reboot with ctrl+alt+del
 
Hmm. I suppose you have already thought about if you have made any changes to the system config? Assuming that it isn't related to any changes you have made perhaps its an issue with the disk controller or the cables? How are the disks connected? Can you change the cable and the port? You should export the pool before changing any ports in case the disk device names change. Or even connect the disks to another system?

Andy.
 
I checked 3 disk on different machines

The problem is with the mount pool to system

I replace all 3 drives

Code: 
zfs umount -a
This will import the pool but will not mount to system!

Code: 
zpool replace ZFS-Vol01 da1 da4
bsd# zpool scrub ZFS-Vol01
bsd# zpool replace ZFS-Vol01 da2 da5
bsd# zpool scrub ZFS-Vol01
bsd# zpool replace-Vol01 da3 da6
bsd# zpool scrub ZFS-Vol01
all successfully

Code: 
bsd# zpool import
bsd# zpool list
NAME        SIZE   USED  AVAIL    CAP  HEALTH  ALTROOT
ZFS-Vol01  1.16T   596G   596G    50%  ONLINE  -
bsd# mount
/dev/da0s1a on / (ufs, local)
devfs on /dev (devfs, local, multilabel)
/dev/da0s1e on /tmp (ufs, local, soft-updates)
/dev/da0s1f on /usr (ufs, local, soft-updates)
/dev/da0s1d on /var (ufs, local, soft-updates)
bsd# zfs mount ZFS-Vol01
shell freezing
or after reboot
Code: 
bsd# zpool import ZFS-Vol01
shell freezing

How can I recover a few files in zfs fs?
 
spamx said:
I checked 3 disk on different machines

The problem is with the mount pool to system

I replace all 3 drives

Code: 
zfs umount -a
This will import the pool but will not mount to system!
Click to expand...

I tested this, [cmd=]zfs umount -a[/cmd] does not import the pool on my system. Another thing that seems very odd from your description is that the import command fails, so I deduce you cannot import the pool, and yet you are able to replace disks on the pool and do a scrub. Doesn't seem to make sense to me...

Also when I suggested testing the drives on another system, I was meaning can you test connecting the drives to another FreeBSD system and do a zpool import?

thanks Andy.
 
Sorry for my english

I did a trick to import pool but not mount pool to system !

without changing / to read-only [cmd=]zpool import ZFS-Vol01[/cmd] command is freezing! I can only reboot with ctrl+alt+del


I remount / to read-only

Code: 
bsd# mount
/dev/da0s1a on / (ufs, local, read-only)
devfs on /dev (devfs, local, multilabel)
/dev/da0s1e on /tmp (ufs, local, soft-updates)
/dev/da0s1f on /usr (ufs, local, soft-updates)
/dev/da0s1d on /var (ufs, local, soft-updates)

Code: 
bsd# zpool import ZFS-Vol01
cannot mount '/ZFS-Vol01': failed to create mountpoint

Code: 
bsd# zfs list
NAME        USED  AVAIL  REFER  MOUNTPOINT
ZFS-Vol01   397G   384G   397G  /ZFS-Vol01

The pool successful imported to system but is not mounted
Now I can exec command [cmd=]zpool replace[/cmd] and [cmd=]zpool scrub[/cmd]

After a successful replace disks and checked the pool

Code: 
bsd# zfs set mountpoint=/var/ZFS-Vol01 ZFS-Vol01
Code: 
bsd# zfs mount ZFS-Vol01
load: 0.01  cmd: zpool 1459 [tx->tx_quiesce_done_cv)] 3.61r 0.00u 0.09s 0% 1908k
load: 0.01  cmd: zpool 1459 [tx->tx_quiesce_done_cv)] 4.25r 0.00u 0.09s 0% 1908k
load: 0.01  cmd: zpool 1459 [tx->tx_quiesce_done_cv)] 4.75r 0.00u 0.09s 0% 1908k
shell freezing
 
You must log in or register to reply here.
Back
Top