Strange problem with mirroring

[Dario_4]

Hello admins,
I've a strange problem during the mirror of disks.
Basically I have 4 disk : 0/6/0 (boot disk), 0/2/0,0/5/0, 0/4/0 and I wuold like to mirror these disks on 1/2/0, 1/5/0, 1/4/0. The boot disk works fine, with the command ftadd.
First of all I create the pv.

pvcreate /dev/rdsk/c1t2d0
pvcreate /dev/rdsk/c1t5d0
pvcreate /dev/rdsk/c1t4d0

But it doesn't create the pv, (also with the options -f).

Someone can suggest me the right way for mirror these disks?

Thanks

[Thierry Poels_1]

Hi,

warning: 1/2/0, 1/5/0, 1/4/0 do NOT necessarily reflect to /dev/rdsk/c1t2d0, /dev/rdsk/c1t5d0 and /dev/rdsk/c1t4d0 !!
Maybe you could use SAM to add the disk to the desired Volume Group.

good luck,
Thierry.

[Dario_4]

Sure,
but from my point of view, it depends from another thing.

root@ [/]
#pvcreate /dev/rdsk/c1t2d0

root@ [/]
#pvcreate /dev/rdsk/c1t4d0
Physical volume "/dev/rdsk/c1t4d0" has been successfully created.

root@ [/]
#pvcreate /dev/rdsk/c1t5d0

root@ [/]
#vgextend /dev/vg01 /dev/dsk/c1t2d0 /dev/dsk/c1t4d0 /dev/dsk/c1t5d0
vgextend: Couldn't open physical volume "/dev/dsk/c1t2d0":
I/O error

[Mark van Hassel]

Hi,

Check the output of ioscan -funCdisk to see what the raw device files are for the hardware paths you specified.

[ian Dennison]

lssf /dev/dsk/c1t2d3 will give the Address of the Disk.

What does a 'diskinfo' tell you?

Share and Enjoy! Ian

[Steven Sim Kok Leong]

Hi,

/dev/dsk/c1t2d0 is usually generated for use by the CD-ROM drive during OS default installation. Please verify from your ioscan output that you are addressing the correct device files for your harddisks.

Hope this helps. Regards.

Steven Sim Kok Leong

[Printaporn_1]

Hi,
Follow this:
Creating Mirrors of Root Logical Volumes
It is highly recommended that you use mirrored root volumes on all
cluster nodes. The following procedure assumes that you are using
separate boot and root volumes; you create a mirror of the boot volume
(/dev/vg00/lvol1), root volume (/dev/vg00/lvol3), and primary swap
(/dev/vg00/lvol2). The procedure cannot be carried out with SAM. In this
example and in the following commands, /dev/dsk/c4t5d0 is the primary
disk and /dev/dsk/c4t6d0 is the mirror; be sure to use the correct device
file names for the root disks on your system.
1. Create a bootable LVM disk to be used for the mirror.
# pvcreate -B /dev/rdsk/c4t6d0
2. Add this disk to the current root volume group.
# vgextend /dev/vg00 /dev/dsk/c4t6d0
3. Make the new disk a boot disk.
# mkboot /dev/rdsk/c4t6d0
4. Copy the correct AUTO file into the new LIF area.
# mkboot -a "hpux -l q (;0)/vmunix" /dev/rdsk/c4t6d0
5. Mirror the boot, root and primary swap logical volumes to the new
bootable disk. Ensure that all devices in vg00, such as /usr, /swap,
etc., are mirrored.
The following is an example of mirroring the boot logical volume:
# lvextend -m 1 /dev/vg00/lvol1 /dev/dsk/c4t6d0
The following is an example of mirroring the root logical volume:
# lvextend -m 1 /dev/vg00/lvol3 /dev/dsk/c4t6d0
The following is an example of mirroring the primary swap logical
volume:
# lvextend -m 1 /dev/vg00/lvol2 /dev/dsk/c4t6d0
6. Update the boot information contained in the BDRA for the mirror
copies of boot, root and primary swap.
# /usr/sbin/lvlnboot -b /dev/vg00/lvol1
# /usr/sbin/lvlnboot -v -r /dev/vg00/lvol3
# /usr/sbin/lvlnboot -s /dev/vg00/lvol2
7. Check if the BDRA is correct.
# /usr/sbin/lvlnboot -R /dev/vg00
8. Verify that the mirrors were properly created.
# lvlnboot -v
The output of this command is shown in a display like the following:
Boot Definitions for Volume Group /dev/vg00:
Physical Volumes belonging in Root Volume Group:
/dev/dsk/c4t5d0 (10/0.5.0) Boot Disk
/dev/dsk/c4t6d0 (10/0.6.0) Boot Disk
Boot: lvol1 on: /dev/dsk/c4t5d0
/dev/dsk/c4t6d0
Root: lvol3 on: /dev/dsk/c4t5d0
/dev/dsk/c4t6d0
Swap: lvol2 on: /dev/dsk/c4t5d0
/dev/dsk/c4t6d0
Dump: lvol2 on: /dev/dsk/c4t6d0, 0

[Michael Tully]

It sounds liks there is a problem with this disk /dev/dsk/c1t2d0

Please post the results of diskinfo /dev/rdsk/c1t2d0
and ioscan -fnkC disk

-Michael

[Stefan Farrelly]

If you cant pvcreate -f a disk then almost always only one of 2 problems;

1. its a hardware problem - your server cant see the disk properly. Check all cables, hardware etc.
2. The disk in question is in use on another server or already on your server - thus even a pvcreate -f will fail. Check your disks very carefully to be sure youre not trying to overwrite a disk in use already.

[Dario_4]

ioscan -funC disk works fine. I see 8 disks.

root [/]
#pvcreate /dev/rdsk/c1t2d0

root [/]
#pvcreate /dev/rdsk/c1t4d0
Physical volume "/dev/rdsk/c1t4d0" has been successfully created.

root@ [/]
#pvcreate /dev/rdsk/c1t5d0

root@TESTBED4 [/]
#vgextend /dev/vg01 /dev/dsk/c1t2d0 /dev/dsk/c1t4d0 /dev/dsk/c1t5d0
vgextend: Couldn't open physical volume "/dev/dsk/c1t2d0":
I/O error



root@ [/]
#pvremove /dev/rdsk/c1t4d0
The physical volume associated with "/dev/rdsk/c1t4d0" has been removed.

root@ [/]
#pvremove /dev/rdsk/c1t2d0
pvremove: Physical Volume "/dev/rdsk/c1t2d0" contains no LVM information

root@ [/]
#pvremove /dev/rdsk/c1t5d0
pvremove: Physical Volume "/dev/rdsk/c1t5d0" contains no LVM information

[Steven Sim Kok Leong]

Hi,

Would you mind showing us the output of ioscan -fnCdisk?

Hope this helps. Regards.

Steven Sim Kok Leong

[Steven Sim Kok Leong]

Hi,

Note that the cdrom also shows up in the ioscan -fnCdisk output.

Just be careful not to mistook the cdrom device file (on many systems it is /dev/dsk/c1t2d0) as a harddisk device file.

Hope this helps. Regards.

Steven Sim Kok Leong

[Dario_4]

#diskinfo /dev/rdsk/c1t2d0
SCSI describe of /dev/rdsk/c1t2d0:
vendor: SEAGATE
product id: ST118273WC
type: direct access
size: 0 Kbytes
bytes per sector: 0


#diskinfo /dev/rdsk/c1t4d0
SCSI describe of /dev/rdsk/c1t4d0:
vendor: SEAGATE
product id: ST118273WC
type: direct access
size: 17783240 Kbytes
bytes per sector: 512


#diskinfo /dev/rdsk/c1t5d0
SCSI describe of /dev/rdsk/c1t5d0:
vendor: SEAGATE
product id: ST118273WC
type: direct access
size: 0 Kbytes
bytes per sector: 0

[Dario_4]

#ioscan -funC disk
Class I H/W Path Driver S/W State H/W Type Description
======================================================================
disk 0 14/0/0.2.0 sdisk CLAIMED DEVICE SEAGATE ST118273WC
/dev/dsk/c0t2d0 /dev/rdsk/c0t2d0
disk 1 14/0/0.4.0 sdisk CLAIMED DEVICE SEAGATE ST118273WC
/dev/dsk/c0t4d0 /dev/rdsk/c0t4d0
disk 2 14/0/0.5.0 sdisk CLAIMED DEVICE SEAGATE ST118273WC
/dev/dsk/c0t5d0 /dev/rdsk/c0t5d0
disk 3 14/0/0.6.0 sdisk CLAIMED DEVICE SEAGATE ST118273WC
/dev/dsk/c0t6d0 /dev/rdsk/c0t6d0
disk 8 14/0/1.2.0 sdisk CLAIMED DEVICE SEAGATE ST118273WC
/dev/dsk/c1t2d0 /dev/rdsk/c1t2d0
disk 9 14/0/1.4.0 sdisk CLAIMED DEVICE SEAGATE ST118273WC
/dev/dsk/c1t4d0 /dev/rdsk/c1t4d0
disk 10 14/0/1.5.0 sdisk CLAIMED DEVICE SEAGATE ST118273WC
/dev/dsk/c1t5d0 /dev/rdsk/c1t5d0
disk 6 14/0/1.6.0 sdisk CLAIMED DEVICE SEAGATE ST118273WC
/dev/dsk/c1t6d0 /dev/rdsk/c1t6d0
disk 4 14/0/2.4.0 sdisk CLAIMED DEVICE TOSHIBA CD-ROM XM-6401TA
/dev/dsk/c2t4d0 /dev/rdsk/c2t4d0

[ian Dennison]

Are you using a Storage Array of some description?

I know the XP512 shows all c?t?d0 Disks as part of a bug in the security software. The 0 Size is a dead giveaway!

Share and Enjoy! Ian

[Steven Sim Kok Leong]

Hi,

The sizes are 0 Kbytes for /dev/rdsk/c1t2d0 and /dev/rdsk/c1t5d0.

Perhaps you would want to try recreating the device files.

# rmsf /dev/rdsk/c1t2d0
# insf -e

Hope this helps. Regards.

Steven Sim Kok Lenog

[Mark van Hassel]

Is STM installed ? If so run the diagnose utility on the two disks that have a 0 size.

[Michael Tully]

I've seen this problem before where
part of an autoraid will create a
device with zero size available.
This will explain whay you can't use
it. The disk could be displaying as
it might just be the spare disk under
RAID 5 for the array.

[Dario_4]

Steven (and alls),
with the command

# rmsf /dev/rdsk/c1t2d0
# insf -e

I see the disk at 0 yet.

#diskinfo /dev/rdsk/c1t5d0
SCSI describe of /dev/rdsk/c1t5d0:
vendor: SEAGATE
product id: ST118273WC
type: direct access
size: 0 Kbytes
bytes per sector: 0

Also for 1.4.0 and 1.5.0

[ian Dennison]

From your last posting, now all 3 disks are showing up as Zero size? Ouch!

This might help the process of elimination

1. Latest SCSI cumulative patch? swlist -l product |grep -i scsi

2. Check SCSI IDs on the Enclosure Units? Swap around cables and try again? What chance for a reboot of the Server?

3. Are you plugging F/W Scsi cables into a 16 bit differential SCSI Board? Is the SCSI Connection on the Main Motherboard or a separate Card? Do you have the most up-to-date drivers for the Card, installed and in the kernel?

Share and Enjoy! Ian

[Dario_4]

Yes,
the SCSI controller is ok.
The patches are:

swlist -l product |grep -i scsi

PHCO_19640 1.0 Manual page for SCSI passthru device
PHKL_14807 1.0 scsi3 cumulative patch
PHKL_17358 1.0 Enable SCSI floppy for 64 bit computers
PHKL_17368 1.0 SCSI Pass thru driver cumulative patch
PHKL_18543 1.0 PM/VM/UFS/async/scsi/io/DMAPI/JFS/perf patch
PHKL_20688 1.0 s700_800 11.00 SCSI IO & Pass-Thru Subsys Cumulative Patch
PHKL_23790 1.0 SCSI Tape (stape) cumulative
SCSI-Passthru B.11.00 HP SCSI-Passthru Driver

[Dario_4]

root@ [/]
#pvcreate /dev/rdsk/c1t2d0



root@ [/]
#pvdisplay -v /dev/dsk/c1t2d0
pvdisplay: Couldn't find the volume group to which
physical volume "/dev/dsk/c1t2d0" belongs.
pvdisplay: Cannot display physical volume "/dev/dsk/c1t2d0".


I don't know why...

[Sandip Ghosh]

If there is no other device on that particular SCSI Controller(14/0/1), try to reset the SCSI BUS for that particular SCSI Controller. Sometime it happens due to SCSI BUS hung problem.

Sandip

[John Palmer]

What does ioscan report as the actual state of the disks.
Please post the output from 'ioscan -fnC disk'.

Your use of the 'u' flag in your earlier post doesn't report the actual hardware state.

Regards,
John

[Steven Sim Kok Leong]

Hi,

What happens when you try to extend your /dev/vg01 after your pvcreate now?

# vgextend /dev/vg01 /dev/dsk/c1t2d0

Hope this helps. Regards.

Steven Sim Kok Leong