File system unmount--make it to mount

[Bunty..]

Hi guys,
My two file systems are unmounted now.
and when i tring to mount..
this is the error:

# mount -F vxfs /dev/vg73/ORACBPD_data_dat01 /ACBP/db/ORACBPD/data/dat01
mount: /ACBP/db/ORACBPD/data/dat01: I/O error


# umount /ACBP/db/ORACBPD/data/idx01
# mount -F vxfs /dev/vg73/ORACBPD_data_idx01 /ACBP/db/ORACBPD/data/idx01
vxfs mount: read of super-block on /dev/vg73/ORACBPD_data_idx01 failed: I/O error



bdf: /ACBP/db/ORACBPD/data/dat01: I/O error

#cd /ACBP/db/ORACBPD/data

# ll
./dat01 not found
total 6
drwxr-xr-x 2 laoracle LAdba 96 Nov 16 08:45 idx01
drwxr-xr-x 3 laoracle LAdba 96 Nov 16 11:19 rbs01
drwxr-xr-x 3 laoracle LAdba 1024 Nov 16 11:19 redoA
drwxr-xr-x 3 laoracle LAdba 1024 Nov 16 11:19 redoB
drwxr-xr-x 3 laoracle LAdba 1024 Nov 16 11:19 redoC
drwxr-xr-x 3 laoracle LAdba 96 Nov 16 11:20 tmp01

# fsck -y /dev/vg73/ORACBPD_data_idx01
vxfs fsck: read of super-block on /dev/vg73/ORACBPD_data_idx01 failed: I/O error
file system check failure, aborting ...


# fsck -m -y /dev/vg73/ORACBPD_data_idx01
vxfs fsck: read of super-block on /dev/vg73/ORACBPD_data_idx01 failed: I/O error

# fsck -F vxfs -m -y /dev/vg73/ORACBPD_data_idx01
vxfs fsck: read of super-block on /dev/vg73/ORACBPD_data_idx01 failed: I/O error

[Bill Hassell · Accepted solution]

You have a failed disk. These are serious errors and cannot be fixed. fsck does not fix disk hardware failures, it only fixes simple directory and disk map mistakes. You'll have to replace the bad disk.

[Sivakumar TS]

Dear Bunty,

I feel the filesystem is currupt. Was there any abrupt shutdown ?

Please do
#lvdisplay -v dev/vg73/ORACBPD_data_idx01

and check for any stale extents.

With Regards,

Siva.

[Devender Khatana]

Hi,

These are symptoms of a failed disk. You should see some STALE PEs in the vgdisplay -v output for this VG. If this VG is not on locally attached disks and is access through SAN than there is a probably a cable fault.

HTH,
Devender

[Steven E. Protter]

Shalom,

When fsck has failed, you are done, disk replacement and data restore is your only option.

You may want to see if you can lvremove the logical volume and vgreduce the disk out of the volume group, just to make the rebuild easier, but it probably will not work.

SEP

[Bunty..]

#lvdisplay -v dev/vg73/ORACBPD_data_idx01
all extents are OK.
is there may be some issue og PEs size.
PEs size is 4MB.
and this LV size is 31GB.

--- Volume groups ---
VG Name /dev/vg73
VG Write Access read/write
VG Status available
Max LV 255
Cur LV 9
Open LV 9
Max PV 16
Cur PV 2
Act PV 2
Max PE per PV 9727
VGDA 4
PE Size (Mbytes) 4
Total PE 19452
Alloc PE 15875
Free PE 3577
Total PVG 0
Total Spare PVs 0
Total Spare PVs in use 0

--- Logical volumes ---
LV Name /dev/vg73/ORACBPD_data_dat01
LV Status available/syncd
LV Size (Mbytes) 31000
Current LE 7750
Allocated PE 7750
Used PV 1

LV Name /dev/vg73/ORACBPD_data_idx01
LV Status available/syncd
LV Size (Mbytes) 11000
Current LE 2750
Allocated PE 2750
Used PV 2

[Devender Khatana]

Hi,

Have you got some messages in the syslog.log while it got unmounted or in dmesg. I suspect a intermediant link failure. Is it on SAN or local disk?

I think it is the best time to check both your disks in that VG for I/O errors through dd command.

#dd if=/dev/rdsk/cxtydz of=/dev/null bs=2048k

Repeat same for other disk in VG. The disk device file is not visible from the posts. so replace /dev/rdsk/cxtydz with that.

HTH,
Devender

[Bunty..]

it is connected to SAN.

#vgdisplay -v /dev/vg73 | more
--- Physical volumes ---
PV Name /dev/dsk/c26t4d1
PV Name /dev/dsk/c29t4d1 Alternate Link
PV Status unavailable
Total PE 9726
Free PE 0
Autoswitch On

PV Name /dev/dsk/c26t4d2
PV Name /dev/dsk/c29t4d2 Alternate Link
PV Status available
Total PE 9726
Free PE 3577
Autoswitch On

how to make it available?


# ioscan -fnC disk | grep c26t4d1
/dev/dsk/c26t4d1 /dev/rdsk/c26t4d1

# ioscan -fnC disk | grep c26t4d1
/dev/dsk/c26t4d1 /dev/rdsk/c26t4d1

# ioscan -fnC disk | grep c26t4d2
/dev/dsk/c26t4d2 /dev/rdsk/c26t4d2

# ioscan -fnC disk | grep c29t4d1
/dev/dsk/c29t4d1 /dev/rdsk/c29t4d1

# ioscan -fnC disk | grep c29t4d2
/dev/dsk/c29t4d2 /dev/rdsk/c29t4d2


should i run this command:
#dd if=/dev/rdsk/c26t4d1 of=/dev/null bs=2048k

will it make PV status abailable

[Devender Khatana]

Hi,

Print the whole information related to /dev/dsk/c26t4d1 & /dev/dsk/c29t4d1.

Also post
#diskinfo -v /dev/rdsk/c26t4d1

&

#diskinfo -v /dev/rdsk/c26t4d1

Allthough this should not happen as you have alternate path to access the disk unless there is some problem at storage/SAN level.

Was there something done/faced in storage at the same time?

If the output of the diskinfo shows the disk as healthy. You can reactivate VG and then the disk should become available if it is really available.

#vgchange -a y /dev/vg73

Then retry mount and if required fsck.

Allthough there may be some corruption to data in such scenarions and you should be able to find that in fsck errors if any.

HTH,
Devender