Boot disk to be changed

[jhingoor]

Hi guys,

our boot disk of one of the server which is in cluster is asking for replacement...May i know the steps involved that i will have to carry out during the downtime....

We do have a mirror copy i.e. another disk which is mirrorred with it...

[Steven E. Protter]

Shalom,

Your best bet is a make_tape_recovery or make_net_recovery prior to replacement.

lvlnboot -v

To check that your mirror is good and the system will boot.

After replacement your system should boot without any issues. You will need to re-mirror the system after disk replacement but that can occur while its up.

The Ignite recovery is insurance in case there is trouble booting off the mirror.

SEP

[jhingoor]

Hi Steven,

We already have taken an ignite backup of entire vg00 for the server...

now how do i restore from this tape....if i need to copy all this data into the new hard disk.........

[Jan van den Ende]

Pardon my ignorance on the HPUX specifics, but is the general concept of MIRROR not to just continue if one of the underlying drives goes away?
Just force the failing/faulty drive out; replace it; and re-form the mirror set from the remaining mirror member and the new drive?
Which would imply NO downtime?

I could supply the specific commands for VMS, or for several NAS & SAN configurations, but alas, I do not have the specific commands for HPUX.
But since this is the whole rationale for mirroring, there MUST be some commands that will allow you to continue operation.

fwiw,

Proost.

Have one on me.

jpe

[James R. Ferguson]

Hi:

You don't provide any details your situation. However, it would be wise to read and follow the following guidelines:

http://www.docs.hp.com/en/5991-1236/When_Good_Disks_Go_Bad.pdf

Regards!

...JRF...

[Robert-Jan Goossens]

Hi,

Take a look at this cookbook.

http://docs.hp.com/en/5991-1236/When_Good_Disks_Go_Bad.pdf

In your case since you are talking about a cluster (serviceguard) I expect you have hot-swappable disks.

# pvchange -a n /dev/dsk/cxtydz

replace the failed disk

# vgcfgrestore â n /dev/vg00 /dev/dsk/cxtydz

# pvchange -a y /dev/dsk/cxtydz

# vgchange -a y vg00

Regards,
Robert-Jan

[freddy_21]

dont forget to create a boot lif after u do mirror

mkboot /dev/dsk/
mkboot -l /dev/rdsk/,cxtydz.
mkboot -a "hpux -lq \ (;0)/stand/vmunix" /dev/dsk/

without boot lif, your disk cant boot up.

[Prashanth.D.S · Accepted solution]

Procedure for disk replacement
----------------------------------
1. Shutdown the server/Npar which is affected by the hard disk failure
2. Remove the faulty disk at path 4/0/10/0/1.0.0
3. After the disk has been replaced, boot using the alternate boot disk at path 4/0/11/0/1.0.0
4. ISL> hpux -lq
5. Use the vgcfgrestore command to restore the LVM configuration data from a configuration backup file to the physical volume:
# vgcfgrestore -n vg00 /dev/rdsk/c1t0d0 (The disk which is replaced)
6. Use mkboot to place boot utilities and an AUTO file in the boot LIF area
# mkboot -l /dev/rdsk/c1t0d0
# mkboot -a "hpux -lq" /dev/rdsk/c1t0d0
Check the contents of the AUTO file with
# lifcp /dev/rdsk/c0t5d0:AUTO -
# lvlnboot -r /dev/vg00/lvol3 /dev/vg00
# lvlnboot -b /dev/vg00/lvol1 /dev/vg00
# lvlnboot -s /dev/vg00/lvol2 /dev/vg00
# lvlnboot -d /dev/vg00/lvol2 /dev/vg00
# lvlnboot -R
# lvlnboot -v
# vgsync vg00
NOTE: The synchronization process can be time consuming, depending on the hardware characteristics and the amount of data


If ODE has to be updated please follow the procedure as follows
List the contents of the LIF directory of the mirror:
# lifls -l /dev/rdsk/c1t0d0
volume ISL10 data size 7984 directory size 8 96/05/28 04:10:13
filename type start size implement created
===============================================================
ISL -12800 584 240 0 96/05/28 04:10:13
AUTO -12289 824 1 0 96/05/28 04:10:13
HPUX -12928 832 848 0 96/05/28 04:10:13
PAD -12290 1680 1652 0 96/05/28 04:10:14
LABEL BIN 3336 8 0 99/01/25 06:00:41
Install the ODE LIF files with mkboot and protect (-p) the existing LIF files which you detected with lifls as described above.
First, check if your system is a pure 64 bit system:
# getconf HW_CPU_SUPP_BITS
For 32 bit and systems that support both 32 and 64 bit:
# mkboot -b /usr/sbin/diag/lif/updatediaglif -p ISL -p AUTO -p HPUX -p PAD -p LABEL /dev/rdsk/c1t0d0
For pure 64 bit systems:
# mkboot -b /usr/sbin/diag/lif/updatediaglif2 -p ISL -p AUTO -p HPUX -p PAD -p LABEL /dev/rdsk/c1t0d0

# lifls -l /dev/rdsk/c1t0d0


One will find additional LIF files.
volume ISL10 data size 7984 directory size 8 96/05/28 04:10:13
filename type start size implement created
===============================================================
ODE -12960 584 880 0 03/05/06 20:27:48
MAPFILE -12277 1464 128 0 03/05/06 20:27:48
SYSLIB -12280 1592 353 0 03/05/06 20:27:48
CONFIGDATA -12278 1952 249 0 03/05/06 20:27:48
SLMOD -12276 2208 133 0 03/05/06 20:27:48
SLDEV -12276 2344 204 0 03/05/06 20:27:48
SLDRIVERS -12276 2552 496 0 03/05/06 20:27:48
SLSCSI -12276 3048 117 0 03/05/06 20:27:48
MAPPER -12279 3168 234 0 03/05/06 20:27:48
IOTEST -12279 3408 208 0 03/05/06 20:27:48
PERFVER -12279 3616 198 0 03/05/06 20:27:48
PVCU -12801 3816 64 0 03/05/06 20:27:48
SSINFO -12286 3880 2 0 03/05/06 20:27:48


Shutdown the system and boot the system from the replaced disk at 4/0/10/0/1.0.0
Check the status.
# vgdisplay -v vg00


NOTE: This is an example...pls make changes to the hardware path and device file names asper system requirement.

Hope this should answer your query.

[Mridul Shrivastava]

If it is mirrored and you know which disk is bad then If you could bring down the server, it would be suggested to replace the disk after bringing down the server and then boot in lvm maintenance mode by interrupting the boot process.

then execute vgcfgrestore -n vg00

then reboot the server and sync vg, I guess this would be easier to complete the task.
But if it is a production server and you can't afford downtime then please go ahead with the steps mentioned by Prashant.

[Torsten.]

Question: What server model and what OS version?

[jhingoor]

hi ,
its rp8420
hp-ux B 11.11

[Torsten.]

If you have problems to find the questionable disk, please ask (and post the hardware path).
Otherwise follow the instructions provided in the document (when_good_disk...) posted at the beginning of this thread.

[jhingoor]

Hi guys,
Our boot disk has been replaced while the server was up..we just had halted the package on it....The steps followed were:


mv lvmtab lvmtab.bck

pvcreate -f /dev/rdsk/c0t6d0


Creating "/etc/lvmtab".
Physical volume "/dev/rdsk/c0t6d0" has been successfully created

mv lvmtab.bck lvmtab

pvchange -a n /dev/dsk/c0t6d0


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/dsk/c0t6d0" has been successfully changed.



vgcfgrestore â n /dev/vg00 /dev/rdsk/c0t6d0

Volume Group configuration has been restored to /dev/rdsk/c0t6d0


pvchange -a y /dev/dsk/c0t6d0

Physical volume "/dev/dsk/c0t6d0" has been successfully changed.


vgchange -a y vg00

Volume Group configuration has been restored to /dev/rdsk/c0t6d0


mkboot -l /dev/dsk/c0t6d0

Creates bootable files in LIF header .

lifls /dev/dsk/c0t6d0

ISL AUTO HPUX PAD LABEL


Just let me know if i have missed something........