topic Re: detailed steps of rx6600 failed root disk replace(11.23) in Operating System - HP-UX

detailed steps of rx6600 failed root disk replace(11.23)

stephen peng — Tue, 16 Nov 2010 09:15:44 GMT

guys,
an rx6600 failed one mirrored disk in vg00 and i am gonna replace it.beside lvreduce, vgreduce, replace failed disk, vgextend, lvextend, is there any other thing I need to handle? i knew that there were three partitions on an 11.23 root disk, can I ignore s1 and s3?

regards
Stephen

Re: detailed steps of rx6600 failed root disk replace(11.23)

rariasn — Tue, 16 Nov 2010 11:12:52 GMT

Hi Stephen:

http://h10025.www1.hp.com/ewfrf/wc/document?docname=c00872150&cc=us&dlc=en&lc=en&jumpid=reg_R1002_USEN

rgs,

Re: detailed steps of rx6600 failed root disk replace(11.23)

Shibin_2 — Tue, 16 Nov 2010 12:54:46 GMT

Hi Stephen,

Read the attached document and follow the steps.

Good luck.

Re: detailed steps of rx6600 failed root disk replace(11.23)

stephen peng — Tue, 16 Nov 2010 13:12:16 GMT

rariasn,
sometimes the disk is just failing but not completely failed, could I consider it failed and follow your method?

regards
Stephen

Re: detailed steps of rx6600 failed root disk replace(11.23)

Torsten. — Tue, 16 Nov 2010 13:13:54 GMT

If this is not a hardware mirror, you need to redirect the new to the old device file after replacement, e.g.

# sasmgr replace_tgt -D /dev/sasd0 -q old_dev=/dev/dsk/c1t4d0 -q new_tgt_hwpath=0/4/1/0.0.0.11.0

(adjust the values to your configuration)

Re: detailed steps of rx6600 failed root disk replace(11.23)

stephen peng — Tue, 16 Nov 2010 13:17:49 GMT

Torsten,
is it necessary to follow your idea at 11.23? i consider REPLACING the disk would reuse the device file.

regards
Stephen

Re: detailed steps of rx6600 failed root disk replace(11.23)

Torsten. — Tue, 16 Nov 2010 13:21:28 GMT

The SAS disk will always get a new device file, so you need to adjust it before you can use the "old" configuration.

Re: detailed steps of rx6600 failed root disk replace(11.23)

stephen peng — Tue, 16 Nov 2010 13:26:13 GMT

Torsten,
can I just use the new device file when
#pvchange -a n
#vgcfgrestore -n vg00
but not using your idea?

regards
Stephen

Re: detailed steps of rx6600 failed root disk replace(11.23)

stephen peng — Tue, 16 Nov 2010 13:27:38 GMT

sorry, it should be
#pvchange -a n s2
#vgcfgrestore -n vg00 s2

Re: detailed steps of rx6600 failed root disk replace(11.23)

Torsten. — Tue, 16 Nov 2010 13:43:42 GMT

No, all your existing configuration contains the old device file.

Re: detailed steps of rx6600 failed root disk replace(11.23)

Torsten. — Tue, 16 Nov 2010 13:47:26 GMT

Get the

HP 8 Internal Port SAS Controller Users Guide

document and follow the chapter

Replacing a Disk Online Using sasmgr replace_tgt or io_redirect_dsf

Re: detailed steps of rx6600 failed root disk replace(11.23)

Torsten. — Tue, 16 Nov 2010 13:53:24 GMT

New link (hard to find since docs.hp.com is dead):

http://h20000.www2.hp.com/bc/docs/support/SupportManual/c02017117/c02017117.pdf

Don't forget to use "idisk" to create the partitions and populate partition 1!

Re: detailed steps of rx6600 failed root disk replace(11.23)

stephen peng — Tue, 16 Nov 2010 14:37:21 GMT

Torsten,
Oh, I never learn this characteristic of SAS disk. From the user guide, new SAS disk would be arranged a new HW ID, so the device file would be differed, right? and rx6600 employs this 8 Internal Port SAS Controller, right?

regards
Stephen

Re: detailed steps of rx6600 failed root disk replace(11.23)

Torsten. — Tue, 16 Nov 2010 14:41:25 GMT

Yes, unless you are using hardware RAID with this default controller or you have an optional P400 smartarray. In this bozh cases you just need to swap the disks - done.

Re: detailed steps of rx6600 failed root disk replace(11.23)

stephen peng — Tue, 16 Nov 2010 15:02:32 GMT

maybe much slower, but I do consider the lvreduce-vgreduce-replace failed disk- vgextend-lvextend way was a safer way when one was not an HP expert!

Re: detailed steps of rx6600 failed root disk replace(11.23)

Anoop Sivan — Wed, 17 Nov 2010 07:04:37 GMT

Hello Stephen,

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/cxtxdx.
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.

# sasmgr get_info -D /dev/sasd0 -q lun=all -q lun_locate

Turn on the diskâ s locator LED to ensure the correct disk is removed.

1. Deactivate the failed disk

# pvchange -a n /dev/dsk/cxtxdxs2

2. Confirm the failed disk (eg: here we use sas bay 1)

# sasmgr get_info -D /dev/sasd0 -q lun=all -q lun_locate
# sasmgr get_info -D /dev/sasd0 -q raid

3. Turn on the LED indicator of the failed disk for replacing

# sasmgr set_attr -D /dev/sasd0 -q lun=/dev/rdsk/cxtxdx -q locate_led=on
# sasmgr get_info -D /dev/sasd0 -q lun=all -q lun_locate

At this point the failed disk is pulled out of the server and 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

4.Note the necessary info, below commands will provide the current disk status.

ioscan -fnC disk

sasmgr get_info -D /dev/sasd0 -q phy=all

sasmgr get_info -D /dev/sasd0 -q raid

sasmgr get_info -D /dev/sasd0 -q lun=all -q lun_locate

5.Associate old device with a new SAS target (whenever a sas disk is replaced the target gets changed)

# sasmgr replace_tgt -D /dev/sasd0 -q old_dev=/dev/dsk/cxtxdx \
-q new_tgt_hwpath=0/4/1/0.0.0.11.0 (change the h/w path for your new disk)

Now you can proceed with the same usual integrity root disk replacement.

create EFI partision-->make s2 partision bootable --> add to vg00 -->restore pvra & vgra....

Note:
=====
If the server has been rebooted for any reason, the special device file is stale and is no longer visible with ioscan, and the
sasmgr replace_tgt will not work unless you have at least B.11.23.03.01 of the driver released January 2007.
Use swlist to check the software driver revision, and upgrade to B.11.23.03.01 or later.

# swlist -l bundle CommonIO
CommonIO B.11.23.03.01 CommonIO

# swlist -l bundle SerialSCSI-00
SerialSCSI-00 B.11.23.03.01 PCI-X SerialSCSI

This software can be obtained free from http://software.hp.com.

>>HP-UX 11i general

o HP-UX Common Bundle
o HP-UX Serial SCSI Driver

Anoop

Re: detailed steps of rx6600 failed root disk replace(11.23)

stephen peng — Wed, 17 Nov 2010 10:58:03 GMT

go to my recent thread please:
http://forums13.itrc.hp.com/service/forums/questionanswer.do?threadId=1456707