Overview
SAS controllers use a different
addressing scheme than parallel SCSI like U320 or SCSI-2.Where SCSI ID and
therefore disk slot numbers were important, SAS uses a unique address of the disks
themselves to identify the disks as part of a LUN (logical unit) or an HP-UX
special device file like /dev/dsk/c0t0d0.
You can take a disk out of one slot (referred to as bays) and put it in
a different slot or bay. The controller
finds it and presents the disk to the O/S with the same special device
file. This does complicate the procedure
needed for replacing a failed disk. !
When a disk fails, and a new disk
is put into the same bay that the failed disk came out of,the SAS controller
knows it is a different disk by its SAS address. The O/S driver assigns the next available
target for the hardware path (viewed with ioscan) and special device file if
insf -e is executed.
If the customer used a legacy
devicefile (e.g. /dev/dsk/c2t2d0) in his LVM configuration, you have to
make sure that the new disk will get the old legacy file again.
This can be done by sasmgr(http://wtec.cup.hp.com/~cpuhw/storage/sas/sasmgr_man.htm
with the “replace_tgt”
option. Even this will result in an Error that the file is still busy, a
consecutive “io_redirect_dsf” on the corresponding persisten file will round up
the move, and result in the old legacy/persisten name pair as before.
If the customer used a persistent
devicefile (e.g. /dev/disk/disk5) in his LVM config, you have to make sure
that the new disk uses the same persistent devicefile again. Ich you
don´t care about the new created legacy dsf and only use the persistent one,
just only use “io_redirect_dsf(1M).
Both methods, sasmgr
replace_tgt and io_redirect_dsf, are
very simple as long as there are no I/O's pending or no I/O drivers that have
the special device file open for reading/writing. This is rarely the case however, and LVM will
continue to try to access that special device file waiting for the failed disk
to return.
You must stop all access to the
special device file first by executing “pvchange –a N” to
deactivate that physical volume.
Once the new disk is inserted,
create the EFI partitions, and vgcfgrestore the LVM information.
As an alternative, you can
unmirror the volume group and vgreduce the bad disk from the volume
group. Following is the procedure using pvchange
and vgcfgrestore to replace a mirrored boot disk in vg00.
1. Check which disk has failed and which devicefiles are
in use
|
# vgdisplay
-v vg00
--- Volume groups ---
VG Name /dev/vg00
VG Write Access read/write
VG Status available
………
--- Logical volumes ---
LV Name /dev/vg00/lvol1
LV Status available/stale
………
LV Name /dev/vg00/lvol8
LV Status available/stale
LV Size (Mbytes) 5696
Current LE 712
Allocated PE 712
Used PV 2
--- Physical volumes ---
PV Name /dev/dsk/disk2_p2
PV Status available
Total PE 4228
Free PE 1224
Autoswitch On
PV Name /dev/dsk/disk3_p2
PV Status unavailable
Total PE 4228
Free PE 1224
Autoswitch On
|
In this case, the disk using the persistent
device file /dev/dsk/disk3_p2 is unavailable.
|
# ioscan -m lun
Class I Lun H/W Path Driver
S/W State H/W Type Health
Description
======================================================================
disk 2 64000/0xfa00/0x0 esdisk
CLAIMED DEVICE online
HP
0/2/1/0.0x5000c50008210fa5.0x0
/dev/disk/disk2
/dev/rdisk/disk2
/dev/disk/disk2_p1
/dev/rdisk/disk2_p1
/dev/disk/disk2_p2
/dev/rdisk/disk2_p2
/dev/disk/disk2_p3 /dev/rdisk/disk2_p3
disk 3
64000/0xfa00/0x1 esdisk NO_HW
DEVICE online HP
0/2/1/0.0x5000c5000820ee41.0x0
/dev/disk/disk3
/dev/rdisk/disk3
/dev/disk/disk3_p1 /dev/rdisk/disk3_p1
/dev/disk/disk3_p2
/dev/rdisk/disk3_p2
/dev/disk/disk3_p3
/dev/rdisk/disk3_p3
|
But also make a note of the failed
legacy device files,. You will need them for sasmgr commands.
|
# ioscan
-fnH 0/2
Class I H/W Path Driver S/W State H/W Type Description
==============================================================================
ba 2 0/2 lba CLAIMED BUS_NEXUS Local PCI-X)
escsi_ctlr 0 0/2/1/0 sasd CLAIMED INTERFACE HP
PCI/PCIr
/dev/sasd0
ext_bus 0 0/2/1/0.0.0 sasd_vbus CLAIMED
INTERFACE SAS Device e
target 0 0/2/1/0.0.0.0 tgt
CLAIMED DEVICE
disk 4 0/2/1/0.0.0.0.0 sdisk
CLAIMED DEVICE HP
D4
/dev/dsk/c0t0d0
/dev/rdsk/c0t0d0
/dev/dsk/c0t0d0s1
/dev/rdsk/c0t0d0s1
/dev/dsk/c0t0d0s2 /dev/rdsk/c0t0d0s2
/dev/dsk/c0t0d0s3
/dev/rdsk/c0t0d0s3
target 1
0/2/1/0.0.0.1 tgt NO_HW DEVICE
disk 0 0/2/1/0.0.0.1.0 sdisk
NO_HW DEVICE HP
D4
/dev/dsk/c0t1d0
/dev/rdsk/c0t1d0
/dev/dsk/c0t1d0s1
/dev/rdsk/c0t1d0s1
/dev/dsk/c0t1d0s2
/dev/rdsk/c0t1d0s2
/dev/dsk/c0t1d0s3 /dev/rdsk/c0t1d0s3
|
Is the disk c0t1d0 alias disk3
an IR Volume or a JBOD ? (look for a disk WWID ending with …..ee41)
|
# sasmgr get_info -D /dev/sasd0 -q raid
Thu Dec 14 14:59:28 2006
---------- PHYSICAL DRIVES
----------
LUN dsf SAS Address Enclosure Bay
Size(MB)
/dev/rdsk/c0t0d0
0x5000c50008210fa5 1 1 140014
/dev/rdsk/c0t1d0 0x5000c5000820ee41 1 2 140014
|
The SAS controller still sees the
disk and has no Raids configured. We have a JBOD disk that, when replaced with
a new one, surely will result in new legacy and persistent devicefiles
|
|
Failed disk
|
New disk
|
|
Type
|
SAS-JBOD
|
SAS-JBOD
|
|
HW Path
|
0/2/1/0.0.0.1.0
|
T.B.D.
|
|
LunPath
|
0/2/1/0.0x5000c5000820ee41.0x0
|
T.B.D.
|
|
Legacy Devicefile
|
/dev/dsk/c0t1d0
|
T.B.D.
|
|
Persistent Devicefile
|
/dev/disk/disk3
|
T.B.D.
|
2. Halt LVM access to the defective disk: (be sure to use
“capital N” with pvchange )
|
# pvchange
-a N /dev/disk/disk3_p2
Warning: Detaching a physical volume reduces the availability of data
within the logical volumes residing on that disk.
Prior to detaching a physical volume or the last available path to it,
verify that there are alternate copies of the data
available on other disks in the volume group.
If necessary, use pvchange(1M) to reverse this operation.
Physical volume
"/dev/disk/disk3_p2" has been successfully changed.
|
3. Turn on the disk's locator LED of disk Bay 2 to ensure the correct
disk is removed:
|
# sasmgr set_attr -D /dev/sasd0 -q lun=/dev/rdisk/disk3 -q
locate_led=on
Locate LED set to ON.
# sasmgr get_info -D /dev/sasd0 -q lun=all -q
lun_locate
/dev/rdsk/c0t0d0 0/2/1/0.0.0.0.0 1 1
OFF
/dev/rdsk/c0t1d0
0/2/1/0.0.0.1.0 1 2
ON
|
4. Replace
the Disk
At this point the disk in bay 2 is
pulled out of the server.
A new disk is inserted in the same
bay. The server should not be rebooted
or taken down between the time the disk fails and the time the new disk is
inserted
Failed disk highlighted in red (State NO_HW)
mirror of that disk in blue (still operating, CLAIMED and untouched)
New replacement disk is in green. (Got a new target ID “2” by SAS
Controller!!!!)
5. Check for new created devicefiles
In this case the customer uses
persistent devicefiles, so we don't care about the new legacy devicefile, but
instead check whick new persistent device the system has created for it:
|
# ioscan -fnH 0/2
Class I
H/W Path Driver S/W State H/W Type Description
==============================================================================
ba 2 0/2 lba CLAIMED BUS_NEXUS Local PCI-X
Bus Adapter (122e)
escsi_ctlr 0 0/2/1/0 sasd CLAIMED INTERFACE HP
PCI/PCI-
X SAS MPT Adapter
/dev/sasd0
ext_bus 0 0/2/1/0.0.0 sasd_vbus CLAIMED
INTERFACE SAS Device I
nterface
target 1
0/2/1/0.0.0.0 tgt CLAIMED DEVICE
disk 1 0/2/1/0.0.0.0.0 sdisk
CLAIMED DEVICE
HP DG
146ABAB4
/dev/dsk/c0t0d0
/dev/rdsk/c0t0d0
/dev/dsk/c0t0d0s1
/dev/rdsk/c0t0d0s1
/dev/dsk/c0t0d0s2
/dev/rdsk/c0t0d0s2
/dev/dsk/c0t0d0s3 /dev/rdsk/c0t0d0s3
target 0
0/2/1/0.0.0.1 tgt NO_HW DEVICE
disk 0 0/2/1/0.0.0.1.0 sdisk
NO_HW DEVICE HP
DG
146ABAB4
/dev/dsk/c0t1d0
/dev/rdsk/c0t1d0
/dev/dsk/c0t1d0s1
/dev/rdsk/c0t1d0s1
/dev/dsk/c0t1d0s2
/dev/rdsk/c0t1d0s2
/dev/dsk/c0t1d0s3
/dev/rdsk/c0t1d0s3
target 1
0/2/1/0.0.2.0 tgt CLAIMED DEVICE
disk 1 0/2/1/0.0.2.0.0 sdisk
CLAIMED DEVICE HP
DG
146ABAB4
/dev/dsk/c0t2d0
/dev/rdsk/c0t2d0
# sasmgr get_info -D /dev/sasd0 -q raid
Thu Dec 14 14:59:28 2006
---------- PHYSICAL DRIVES
----------
LUN dsf SAS Address Enclosure Bay
Size(MB)
/dev/rdsk/c0t0d0
0x5000c50008210fa5 1 1 140014
/dev/rdsk/c0t2d0 0x5000cca000101799 1
2 140014
|
|
# ioscan -m dsf
Persistent DSF Legacy DSF(s)
========================================
/dev/rdisk/disk2
/dev/rdsk/c0t0d0
/dev/rdisk/disk2_p1
/dev/rdsk/c0t0d0s1
/dev/rdisk/disk2_p2
/dev/rdsk/c0t0d0s2
/dev/rdisk/disk2_p3
/dev/rdsk/c0t0d0s3
/dev/rdisk/disk3
/dev/rdsk/c0t1d0
/dev/rdisk/disk3_p1
/dev/rdsk/c0t1d0s1
/dev/rdisk/disk3_p2
/dev/rdsk/c0t1d0s2
/dev/rdisk/disk3_p3
/dev/rdsk/c0t1d0s3
/dev/rdisk/disk5
/dev/rdsk/c0t2d0
<- new disk !
# ioscan -m lun
Class I
Lun H/W Path Driver S/W State
H/W Type Health Description
======================================================================
disk 2 64000/0xfa00/0x0 esdisk
CLAIMED DEVICE online
HP
0/2/1/0.0x5000c50008210fa5.0x0
/dev/disk/disk2
/dev/rdisk/disk2
/dev/disk/disk2_p1
/dev/rdisk/disk2_p1
/dev/disk/disk2_p2
/dev/rdisk/disk2_p2
/dev/disk/disk2_p3 /dev/rdisk/disk2_p3
disk 3
64000/0xfa00/0x1 esdisk NO_HW
DEVICE online HP
0/2/1/0.0x5000c5000820ee41.0x0
/dev/disk/disk3
/dev/rdisk/disk3
/dev/disk/disk3_p1 /dev/rdisk/disk3_p1
/dev/disk/disk3_p2
/dev/rdisk/disk3_p2
/dev/disk/disk3_p3
/dev/rdisk/disk3_p3
disk 5
64000/0xfa00/0x2 esdisk CLAIMED
DEVICE online HP
0/2/1/0. 0x5000cca000101799.0x0
/dev/disk/disk5
/dev/rdisk/disk5
# diskinfo /dev/disk/disk5
diskinfo: Character device required
# diskinfo /dev/rdisk/disk3
SCSI describe of /dev/rdisk/disk3:
vendor: HP
product id: DG146ABAB4
type: direct access
size: 143374744 Kbytes
bytes per sector: 512
|
Collect your data now :
|
|
Failed disk
|
New disk
|
|
Type
|
SAS-JBOD
|
SAS-JBOD
|
|
HW Path
|
0/2/1/0.0.0.1.0
|
0/2/1/0.0.0.2.0
|
|
LunPath
|
0/2/1/0.0x5000c5000820ee41.0x0
|
0/2/1/0.
0x5000cca000101799.0x0
|
|
Legacy
Devicefile
|
/dev/dsk/c0t1d0
|
/dev/dsk/c0t2d0
|
|
Persistent
Devicefile
|
/dev/disk/disk3
|
/dev/disk/disk5
|
6. Stop the LED (now using the new legacy devicefile)
|
# sasmgr get_info -D /dev/sasd0 -q lun=all -q lun_locate
/dev/rdsk/c0t0d0 0/2/1/0.0.0.0.0 1 1
OFF
/dev/rdsk/c0t2d0 0/2/1/0.0.0.2.0 1 2
ON
# sasmgr set_attr -D /dev/sasd0 -q lun=/dev/rdsk/c0t2d0 -q locate_led=off
|
7. Restore
the IA64 Partitioning Scheme of the new boot disk
Note: As the tools to retain the old defvcefiles after disk
replacement do only allow replacement of disks with identical number of
devicefiles, You have to make sure now that the new disk has the same partitioning
scheme as the failed one.
If you would try to move the disk5 to the old disk3 name, you will get an Error
Message:
Example:
# io_redirect_dsf -d /dev/rdisk/disk3 -n /dev/rdisk/disk5
Number of old DSFs=8.
Number of new DSFs=2.
The number of old and new DSFs must be the same.
Be aware that you use the new created devicefiles at
this time.
- Create a partition description file:
|
# vi /tmp/partitionfile
3
EFI 500MB
HPUX 100%
HPSP 400MB
|
Use idisk(1M) command to partition the disk according to
this file: \
|
# idisk -wf /tmp/partitionfile /dev/rdisk/disk5
idisk version: 1.2
********************** WARNING ***********************
If you continue you may destroy all data on this disk.
Do you wish to continue(yes/no)? yes
|
Create the new device files for the new partitions
(disk3_p1,_p2_p3)
|
# insf -e –Cdisk
|
|
# ioscan
-m lun
Class I
Lun H/W Path Driver S/W State
H/W Type Health Description
======================================================================
disk 2 64000/0xfa00/0x0 esdisk
CLAIMED DEVICE online
HP
0/2/1/0.0x5000c50008210fa5.0x0
/dev/disk/disk2
/dev/rdisk/disk2
/dev/disk/disk2_p1
/dev/rdisk/disk2_p1
/dev/disk/disk2_p2 /dev/rdisk/disk2_p2
/dev/disk/disk2_p3
/dev/rdisk/disk2_p3
disk 3
64000/0xfa00/0x1 esdisk NO_HW
DEVICE online HP
0/2/1/0.0x5000c5000820ee41.0x0
/dev/disk/disk3 /dev/rdisk/disk3
/dev/disk/disk3_p1
/dev/rdisk/disk3_p1
/dev/disk/disk3_p2
/dev/rdisk/disk3_p2
/dev/disk/disk3_p3
/dev/rdisk/disk3_p3
disk 5
64000/0xfa00/0x2 esdisk CLAIMED
DEVICE online HP
0/2/1/0. 0x5000cca000101799.0x0
/dev/disk/disk5
/dev/rdisk/disk5
/dev/disk/disk5_p1
/dev/rdisk/disk5_p1
/dev/disk/disk5_p2 /dev/rdisk/disk5_p2
/dev/disk/disk5_p3
/dev/rdisk/disk5_p3
# ioscan -fnH 0/2
Class I
H/W Path Driver S/W State H/W Type Description
==============================================================================
ba 2
0/2 lba CLAIMED BUS_NEXUS Local PCI-X
escsi_ctlr 0 0/2/1/0 sasd CLAIMED INTERFACE HP
PCI/PCI-X SAS MPT Adapter
/dev/sasd0
ext_bus 0 0/2/1/0.0.0 sasd_vbus CLAIMED
INTERFACE SAS Device I
nterface
target 1
0/2/1/0.0.0.0 tgt CLAIMED DEVICE
disk 1 0/2/1/0.0.0.0.0 sdisk
CLAIMED DEVICE HP
DG
146ABAB4
/dev/dsk/c0t0d0 /dev/rdsk/c0t0d0
/dev/dsk/c0t0d0s1
/dev/rdsk/c0t0d0s1
/dev/dsk/c0t0d0s2
/dev/rdsk/c0t0d0s2
/dev/dsk/c0t0d0s3
/dev/rdsk/c0t0d0s3
target 0
0/2/1/0.0.0.1 tgt
NO_HW DEVICE
disk 0 0/2/1/0.0.0.1.0 sdisk
NO_HW DEVICE HP
DG
146ABAB4
/dev/dsk/c0t1d0
/dev/rdsk/c0t1d0
/dev/dsk/c0t1d0s1 /dev/rdsk/c0t1d0s1
/dev/dsk/c0t1d0s2
/dev/rdsk/c0t1d0s2
/dev/dsk/c0t1d0s3
/dev/rdsk/c0t1d0s3
target 1
0/2/1/0.0.2.0 tgt CLAIMED DEVICE
disk 1 0/2/1/0.0.2.0.0 sdisk CLAIMED DEVICE HP
DG
146ABAB4
/dev/dsk/c0t2d0
/dev/rdsk/c0t2d0
/dev/dsk/c0t2d0s1
/dev/rdsk/c0t2d0s1
/dev/dsk/c0t2d0s2
/dev/rdsk/c0t2d0s2
/dev/dsk/c0t2d0s3
/dev/rdsk/c0t2d0s3
# ioscan -m dsf
Persistent DSF Legacy DSF(s)
========================================
/dev/rdisk/disk2
/dev/rdsk/c0t0d0
/dev/rdisk/disk2_p1
/dev/rdsk/c0t0d0s1
/dev/rdisk/disk2_p2
/dev/rdsk/c0t0d0s2
/dev/rdisk/disk2_p3
/dev/rdsk/c0t0d0s3
/dev/rdisk/disk3 /dev/rdsk/ c0t1d0 <-
failed disk !
/dev/rdisk/disk3_p1
/dev/rdsk/c0t1d0s1
/dev/rdisk/disk3_p2
/dev/rdsk/c0t1d0s2
/dev/rdisk/disk3_p3
/dev/rdsk/c0t1d0s3
/dev/rdisk/disk5 /dev/rdsk/c0t2d0 <-
new disk !
/dev/rdisk/disk5_p1
/dev/rdsk/c0t2d0s1
/dev/rdisk/disk5_p2
/dev/rdsk/c0t2d0s2
/dev/rdisk/disk5_p3
/dev/rdsk/c0t2d0s3
|
1. Redirect the IO from the new Device to the old Devicefile for
the legacy devicefile:
|
# sasmgr replace_tgt -D /dev/sasd0 -q old_dev=/dev/dsk/c0t1d0
–q new_tgt_hwpath=0/2/1/0.0.0.2.0
WARNING: This is a DESTRUCTIVE operation.
No CRA is done by the driver for this command.
Please ensure that there are no outstanding I/O requests
on the old_dev and the LUN's attached to the
new_tgt_hwpath before proceeding with this command.
Do you want to continue ?(y/n) [n]...
ERROR: Unable to remove mapping of the persistent LUN with the legacy dsf.
Possibly the LUN is open. Please execute scsimgr replace_leg_dsf command
Please see the scsimgr(1M) manpage for more details.
|
What happened ? Since there is
still an active mapping from “c0t2d0” to the persistent dsf “disk5”, the system
complains that you renamed one part of it. But, it alteady deleted the new
legacy dsf´s:
|
# ioscan -fnH 0/2
Class I
H/W Path Driver S/W State H/W Type Description
==============================================================================
ba 2
0/2 lba CLAIMED BUS_NEXUS Local PCI-X
escsi_ctlr 0 0/2/1/0 sasd CLAIMED INTERFACE HP
PCI/PCI-X SAS MPT Adapter
/dev/sasd0
ext_bus 0 0/2/1/0.0.0 sasd_vbus CLAIMED
INTERFACE SAS Device I
nterface
target 1
0/2/1/0.0.0.0 tgt CLAIMED DEVICE
disk 1 0/2/1/0.0.0.0.0 sdisk
CLAIMED DEVICE HP
DG
146ABAB4
/dev/dsk/c0t0d0 /dev/rdsk/c0t0d0
/dev/dsk/c0t0d0s1
/dev/rdsk/c0t0d0s1
/dev/dsk/c0t0d0s2
/dev/rdsk/c0t0d0s2
/dev/dsk/c0t0d0s3
/dev/rdsk/c0t0d0s3
target 0
0/2/1/0.0.0.1 tgt NO_HW DEVICE
disk 0 0/2/1/0.0.0.1.0 sdisk
NO_HW DEVICE HP
DG
146ABAB4
/dev/dsk/c0t1d0
/dev/rdsk/c0t1d0
/dev/dsk/c0t1d0s1
/dev/rdsk/c0t1d0s1
/dev/dsk/c0t1d0s2
/dev/rdsk/c0t1d0s2
/dev/dsk/c0t1d0s3
/dev/rdsk/c0t1d0s3
target 1
0/2/1/0.0.2.0 tgt NO_HW DEVICE deleted !
|
Now, let´s re-animate the persisten devicefile also:
for the persisten devicefile:
|
# io_redirect_dsf –d /dev/disk/disk3 –n
/dev/disk/disk5
|
- Verify it:
|
# ioscan -m lun
Class I Lun H/W Path Driver
S/W State H/W Type Health
Description
======================================================================
disk 2 64000/0xfa00/0x0 esdisk
CLAIMED DEVICE online
HP
0/2/1/0.0x5000c50008210fa5.0x0
/dev/disk/disk2 /dev/rdisk/disk2
/dev/disk/disk2_p1
/dev/rdisk/disk2_p1
/dev/disk/disk2_p2
/dev/rdisk/disk2_p2
/dev/disk/disk2_p3
/dev/rdisk/disk2_p3
disk 5
64000/0xfa00/0x2 esdisk CLAIMED
DEVICE online
HP
0/2/1/0. 0x5000cca000101799.0x0
/dev/disk/disk3
/dev/rdisk/disk3
/dev/disk/disk3_p1
/dev/rdisk/disk3_p1
/dev/disk/disk3_p2
/dev/rdisk/disk3_p2
/dev/disk/disk3_p3 /dev/rdisk/disk3_p3
# ioscan -fnH 0/2
Class I
H/W Path Driver S/W State H/W Type Description
==============================================================================
ba 2 0/2 lba CLAIMED BUS_NEXUS Local PCI-X
escsi_ctlr 0 0/2/1/0 sasd CLAIMED INTERFACE HP
PCI/PCI-X SAS MPT Adapter
/dev/sasd0
ext_bus 0 0/2/1/0.0.0 sasd_vbus CLAIMED
INTERFACE SAS Device I
nterface
target 1 0/2/1/0.0.0.0 tgt
CLAIMED DEVICE
disk 1 0/2/1/0.0.0.0.0 sdisk
CLAIMED DEVICE HP
DG
146ABAB4
/dev/dsk/c0t0d0
/dev/rdsk/c0t0d0
/dev/dsk/c0t0d0s1
/dev/rdsk/c0t0d0s1
/dev/dsk/c0t0d0s2
/dev/rdsk/c0t0d0s2
/dev/dsk/c0t0d0s3
/dev/rdsk/c0t0d0s3
target 0
0/2/1/0.0.0.1 tgt CLAIMED
DEVICE
disk 0 0/2/1/0.0.0.1.0 sdisk
CLAIMED DEVICE HP
DG
146ABAB4
/dev/dsk/c0t1d0
/dev/rdsk/c0t1d0
/dev/dsk/c0t1d0s1
/dev/rdsk/c0t1d0s1
/dev/dsk/c0t1d0s2
/dev/rdsk/c0t1d0s2
/dev/dsk/c0t1d0s3
/dev/rdsk/c0t1d0s3
target 1
0/2/1/0.0.2.0 tgt NO_HW DEVICE
deleted !
# ioscan -m dsf
Persistent DSF Legacy DSF(s)
========================================
/dev/rdisk/disk2
/dev/rdsk/c0t0d0
/dev/rdisk/disk2_p1
/dev/rdsk/c0t0d0s1
/dev/rdisk/disk2_p2
/dev/rdsk/c0t0d0s2
/dev/rdisk/disk2_p3
/dev/rdsk/c0t0d0s3
/dev/rdisk/disk3 /dev/rdsk/ c0t1d0
/dev/rdisk/disk3_p1
/dev/rdsk/c0t1d0s1
/dev/rdisk/disk3_p2
/dev/rdsk/c0t1d0s2
/dev/rdisk/disk3_p3
/dev/rdsk/c0t1d0s3
|
Bingo ! The old dsf´s
“c0t0d0 and “disk3” are operational again !
9.
Initialize the EFI FAT Partition and fill boot areas:
Use efi_fsinit(1M) to initialize the FAT filesystem on the
EFI partition:
|
# efi_fsinit -d /dev/rdisk/disk3_p1
|
NOTE:
This step is not neccessary if it can be guaranteed that the mirror disk does
not contain a valid EFI filesystem. In this case efi_fsinit(1M) will be done
automatically by the subsequent mkboot(1M) command. But if you take e.g. an old
UX 11.22 boot disk as mirror disk, mkboot will not automatically run
efi_fsinit. As a result only 100MB of the 500MB EFI partition (s1) can be used.
Use mkboot(1M) to format the EFI
partition (s1) and populate it with the EFI files below /usr/lib/efi/ and to format the LIF
volume (part of s2) and populate it with the LIF files (ISL, AUTO, HPUX, LABEL)
below /usr/lib/uxbootlf:
|
# mkboot -e -l /dev/rdisk/disk3
# efi_ls -d /dev/rdisk/disk3_p1 (to
check EFI)
FileName Last Modified Size
EFI/ 11/ 5/2003 0
STARTUP.NSH 11/ 5/2003 296
total space 523251712 bytes,
free space 520073216 bytes
# lifls -l /dev/rdisk/disk3_p2 (to
check LIF)
volume ISL10 data size 7984 directory size 8 07/01/10 17:13:32
filename type start
size implement created
===============================================================
ISL -12800 584 242
0 07/01/10 17:13:32
AUTO -12289 832 1
0 07/01/10 17:13:32
HPUX -12928 840 1024
0 07/01/10 17:13:32
PAD -12290 1864 1468
0 07/01/10 17:13:33
LABEL BIN 3336 8
0 08/08/26 20:07:24
|
Check the content of AUTO file on the original Bootdisks EFI
partition:
|
# efi_cp -d /dev/rdisk/disk2_p1 -u
/EFI/HPUX/AUTO /tmp/x
# cat /tmp/x
boot vmunix
|
NOTE:
Specify the -lq option if prefer that your system boots up without
interruption in case of a disk failure:
on
the original boot disk:
|
# mkboot -a "boot vmunix -lq"
/dev/rdisk/disk2
|
and
on the replaced new mirror disk:
|
# mkboot -a "boot vmunix -lq"
/dev/rdisk/disk3
|
Copy the HP service partition (skip this, if you don't have
a service partition)
|
# dd if=/dev/rdisk/disk2_p3
of=/dev/rdisk/disk3_p3 bs=1024k
|
10. Restore LVM Configuration
Now the new disk is
partitioned and equipped with boot headers, you can restore the LVM data to the
OS partition “p2”:
|
# vgcfgrestore -n /dev/vg00 /dev/rdisk/disk3_p2
|
Restore
LVM access to the disk.
If you did not reboot the system
in Step 2http://docs.hp.com/en/5992-3385/ch04s08.html, reattach the disk as follows:
|
# pvchange –a y /dev/disk/disk3_p2
|
If you did reboot the system, reattach
the disk by reactivating the volume group as follows:
# vgchange -a y /dev/vg00
|
|
|
|
|
NOTE: The vgchange command with the -a y option can be run on a
volume group that is deactivated or already activated. It attaches all paths
for all disks in the volume group and resumes automatically recovering any
disks in the volume group that had been offline or any disks in the volume
group that were replaced. Therefore, run vgchange only after all work
has been completed on all disks and paths in the volume group, and it is
necessary to attach them all.
|
Syncronize volume group data (only if sync does not start
automatically):
|
# cd /tmp
# nohup vgsync /dev/vg00 & (output
see /tmp/nohup.out)
|
Initialize/check boot information on the disk.
Check if content of LABEL file
(i.e. root, boot, swap and dump device definition) has been initialized (done
by lvextend) on the mirror disk:
|
# lvlnboot -v
Boot Definitions for Volume Group /dev/vg00:
Physical Volumes belonging in Root Volume Group:
/dev/disk/disk2_p2 -- Boot Disk
/dev/disk/disk3_p2 -- Boot Disk
Boot: lvol1 on: /dev/disk/disk2_p2
/dev/disk/disk3_p2
Root: lvol3 on: /dev/disk/disk2_p2
/dev/disk/disk3_p2
Swap: lvol2 on: /dev/disk/disk2_p2
/dev/disk/disk3_p2
Dump: lvol2 on: /dev/disk/disk2_p2, 0
|
If not, then set it:
|
# lvlnboot -r /dev/vg00/lvol3
# lvlnboot -b /dev/vg00/lvol1
# lvlnboot -s /dev/vg00/lvol2
# lvlnboot -d /dev/vg00/lvol2
|
update EFI Boot Configuration to recognize the new Disk ID:
|
# # setboot
Primary bootpath :
0/2/1/0.0x5000c50008210fa5.0x0 (/dev/rdisk/disk2)
HA Alternate bootpath : 0/1/1/1 (LAN Interface)
Alternate bootpath : 0/1/1/1 (LAN
Interface)
Autoboot is ON (enabled)
Hyperthreading : ON
: ON (next boot)
# setboot –h /dev/disk/disk3 (HA-alternate)
# setboot –a /dev/disk/disk3 (alternate
)
# setboot (to check it)
# setboot
Primary bootpath :
0/2/1/0.0x5000c50008210fa5.0x0 (/dev/rdisk/disk2)
HA Alternate bootpath :
0/2/1/0. 0x5000cca000101799.0x0 (/dev/rdisk/disk3)
Alternate bootpath : 0/2/1/0.
0x5000cca000101799.0x0 (/dev/rdisk/disk3)
Autoboot is ON (enabled)
Hyperthreading : ON
: ON (next boot)
|
This
will write new Entries into the EFI Boot Menu in the order:
- Primary Path
- HAAlternate Path
- Alternate Path
