failed boot HP9000/K460

[Mike_Ca Li]

I'm experiencing OS bootup issue after I replaced a SCSI disk similar to old thread (servers/9000 section) is "failed boot HP9000/D350"
Originaly I had 4 X 9 GB with IDs 3, 4, 5, 6
I had /dev/vg00 on /dev/dsk/c0t1d0 and /dev/dsk/c0t2d0 (mirrored)
and vg01 on /dev/dsk/c0t3d0 and /dev/dsk/c0t4d0
I broke the mirror for vg01 and vg01 only had c0t3d0.
I want to replace c0t4d0 with a 36 GB disk

I disconnected disks with IDs 5 and 6 and replaced with new SCSI HD ID 6.
Getting the WARN 80F5 error and can't find the boot device (cannot find ENTRY_TEST).
I did a SEA to search for a boot path:
P0 10/0.8 fwscsi.8 Random access media
P1 10/0.4 fwscsi.4 "
P2 10/0.3 fwscsi.3 "
P3 10/12/5.0 sescsi.0 Sequential

Pri is set to 10/0.6 , my question is that did I replace/disconnected the boot devices?
Anyone has idea how SCSI Id is related to fwscsi.## ? Thank you

[Patrick Wallek · Accepted solution]

I would say yes, there is a good change you replaced a boot device.

The information you gave contradicts itself. You say fir that you had disks with SCSI ID's 3,4,5,6. Then you say vg00 had disks c0t1d0 and c0t2d0. That is a contradiction. It's impossible. The 't' portion of the device file is the scsi id. t1 and t2 indicates scsi id's 1 and 2, which it looks like you don't have.

I would put the disks from id 5 and 6 back. Those are more than likely your vg00 disks.

Once the system boots again, do a 'vgdisplay -v vg00' and look VERY CAREFULLY at your disk device files included in the VG.

[Steven E. Protter]

If you mirrored the boot disk prior to the failure, you should be okay.

ioscan -fnC disk

SEP

[Mike_Ca Li]

You are right on. /dev/rdsk/c0t6d0 and c0t5d0 were the vg00 and I am able to boot up now. But now I can't see vg01 even tough I connected back my disks. ioscan -fnC disk
displays all the 4 disks. Thank you

[Vincente Fernandes]

Run this command the output should be like
#strings /etc/lvmtab
/dev/vg01
/dev/dsk/cxtxdx
/dev/dsk/cxtxdx
Does the above cxtxdx match with ioscan output? What happens with "vgchange -a y /dev/vg01"?
If not you can always import the vg01 volume group
If you have the mapfile import vg01 using mapfile using below 1 of the syntax.
/usr/sbin/vgimport [-m mapfile] [-p] [-v] [-f infile] vg_name pv_path
/usr/sbin/vgimport -m mapfile -s -v vg_name

[Pete Randall]

Does the directory /dev/vg01 exist? Does it contain the /dev/vg01/group file? If not create both:

mkdir /dev/vg01
mknod /dev/vg01/group c64 0x010000

In either case, then import vg01

vgimport /dev/vg01

and activate it:

vgchange -a y /dev/vg01


Pete

[Mike_Ca Li]

Thanks Vince and Peter:
scledibk:/ {130}> cd /dev
scledibk:/dev {131}> ls -l vg01
total 0
cr--r--r-- 1 root sys 64 0x010000 Nov 25 2002 group
br--r----- 1 root sys 64 0x010001 Nov 25 2002 lvol1
cr--r----- 1 root sys 64 0x010001 Nov 25 2002 rlvol1
scledibk:/dev {132}> more vg01/group
scledibk:/dev {133}> vgimport /dev/vg01
vgimport: Volume group "/dev/vg01" already exists in the "/etc/lvmtab" file.
scledibk:/dev {134}> vgchange -a y /dev/vg01
vgchange: Warning: Couldn't attach to the volume group physical volume "/dev/dsk
/c0t3d0":
The path of the physical volume refers to a device that does not
exist, or is not configured into the kernel.
vgchange: Warning: couldn't query physical volume "/dev/dsk/c0t3d0":
The specified path does not correspond to physical volume attached to
this volume group
vgchange: Warning: couldn't query all of the physical volumes.
vgchange: Couldn't activate volume group "/dev/vg01":
Quorum not present, or some physical volume(s) are missing.

[Kent Ostby]

Mike -- It can't talk to the disk at ID 3 .. (c0t3d0) .. if this was disconnected then you know why.

If not then it might be bad.

Do a :

diskinfo /dev/rdsk/c0t3d0 and see what the size is (hopefully non zero).

Also, try to dd from that disk:

dd if=/dev/rdsk/c0t3d0 of=/dev/null bs=64k

This will test the disk at a HW level instead of an LVM level.

Best regards,

Kent M. Ostby

[Mike_Ca Li]

I had vg01 and had 1 logical vol /dev/vg01/lvol1 which is mirrored with
/dev/dsk/c0t3d0 and /dev/dsk/c0t4d0

I did a
lvreduce -m 0 /dev/vg01/lvol1
vgreduce vg01 /dev/dsk/c0t4d0

I took out c0t3d0 instead of c0t4d0 from the disk bay.
so c0t4d0 is still connected and can be detected by OS.
Class I H/W Path Driver S/W State H/W Type Description
=============================================================
disk 6 10/0.2.0 sdisk CLAIMED DEVICE SEAGATE ST336704LC
/dev/dsk/c0t2d0 /dev/rdsk/c0t2d0
disk 1 10/0.4.0 sdisk CLAIMED DEVICE SEAGATE ST19171W
/dev/dsk/c0t4d0 /dev/rdsk/c0t4d0
disk 2 10/0.5.0 sdisk CLAIMED DEVICE SEAGATE ST19171W
/dev/dsk/c0t5d0 /dev/rdsk/c0t5d0
disk 3 10/0.6.0 sdisk CLAIMED DEVICE SEAGATE ST19171W
/dev/dsk/c0t6d0 /dev/rdsk/c0t6d0

Can I get the data for /dev/vg01/lvol1 from c0t4d0?
If so what are the steps? or do I need to put back c0t3d0?

[Devender Khatana]

Hi,

As you reduced it from this device then reduced VG from the device the information there can not be accessed easily.
When you have another copy of this in good form put that back and access data.

HTH,
Devender

[Mike_Ca Li]

Is the data which was mirror to c0t4d0 still available?

[Devender Khatana]

Hi,

Yes,data is still there in drive as you did not overwrite anything to drive after removing it from VG. But it will not be accessed straight away as your remove this disk from VG and kept other disk as active. So it is advised to take data having plug in the other disk. Recovering it from this disk will be a slightly a long process.
Is there any problems in accessing the second disk?

HTH,
Devender

[Mike_Ca Li]

I'm just interested on how to get to the data on c0t4d0. Thank you. I'll assign points.

[Devender Khatana]

Hi,

This even I never tried but you can try by doing something like following -

Your vg01 should not be accessible while doing this as it belongs to that only. Just see "strings /etc/lvmtab". If it displays something belonging to /dev/vg01 then it would be /dev/dsk/c0t3d0. Just do a vgscan and see"strings /etc/lvmtab" again. Now it should not display /dev/vg01.

Now you try this.

#cd /dev/vg01
#vgimport /dev/vg01 /dev/dsk/c0t4d1
#vgchange -a y /dev/vg01 ( If gives error then use )

#vgchange -a y -q n /dev/vg01
#mount /dev/vg01/lvol1 /mountpoint

Do it and let us know if you face any problems.

HTH,
Devender

[Mike_Ca Li]

server:/dev {135}>
server:/dev {300}> cd /dev/vg01
server:/dev/vg01 {301}> vgimport /dev/vg01 /dev/dsk/c0t4d1
vgimport: Volume group "/dev/vg01" already exists in the "/etc/lvmtab" file.
server:/dev/vg01 {303}> vgchange -a y /dev/vg01
Volume group "/dev/vg01" has been successfully changed.
server:/dev/vg01 {304}> mount /dev/vg01/lvol1 /apps2
/dev/vg01/lvol1: No such device or address

I can't also create Filesystem on lvol1.
Thanks.

[Mike_Ca Li]

server:/dev/vg01 {308}> vgscan
vgscan: The physical volume "/dev/dsk/c0t2d0" is already recorded in the "/etc/l
vmtab" file.
vgscan: The physical volume "/dev/dsk/c0t4d0" is already recorded in the "/etc/l
vmtab" file.
vgscan: The physical volume "/dev/dsk/c0t5d0" is already recorded in the "/etc/l
vmtab" file.
vgscan: The physical volume "/dev/dsk/c0t6d0" is already recorded in the "/etc/l
vmtab" file.

server:/dev/vg01 {309}> strings /etc/lvmtab
/dev/vg00
/dev/dsk/c0t5d0
/dev/dsk/c0t6d0
/dev/vg01
/dev/dsk/c0t4d0
/dev/vg02
/dev/dsk/c0t2d0

server:/dev/vg01 {310}> vgdisplay -v /dev/vg01
--- Volume groups ---
VG Name /dev/vg01
VG Write Access read/write
VG Status available
Max LV 255
Cur LV 0
Open LV 1
Max PV 16
Cur PV 1
Act PV 1
Max PE per PV 2169
VGDA 2
PE Size (Mbytes) 4
Total PE 2169
Alloc PE 0
Free PE 2169
Total PVG 0
Total Spare PVs 0
Total Spare PVs in use 0

--- Logical volumes ---
LV Name /dev/vg01/lvol1
LV Status available/syncd
LV Size (Mbytes) 0
Current LE 0
Allocated PE 0
Used PV 0


--- Physical volumes ---
PV Name /dev/dsk/c0t4d0
PV Status available
Total PE 2169
Free PE 2169
Autoswitch On

server:/dev/vg01 {312}> lvcreate -L 8676 /dev/vg01
lvcreate: Logical volume "/dev/vg01/lvol1" already exists.
server:/dev/vg01 {314}> newfs -F vxfs /dev/vg01/rlvol1

vxfs newfs: Cannot open /dev/vg01/rlvol1: No such device or address

Pls point out what could be wrong. Thank you

[Devender Khatana]

Hi,

A lvol in your vg with name lvol1 is already existing allthough a device file do not exist for that. You can try listing the contents on /dev/vg01 directory and try doing it by specifying a different lvname with -n option of lvcreate. Man lvcreate for details.

HTH,
Devender

[Kent Ostby]

I suspect that /dev/vg01/lvol1 exists but /dev/vg01/rlvol1 doesn't.

Check for that.

If it doesn't then recreate it. Probably with:

mknod /dev/vg01/rlvol1 c 0x64 0x010001

[Mike_Ca Li]

Hi Kent:
It does exist. I created lvol2, how to take out lvol1 or rvol1? Thks.
server:/dev/vg01 {339}> ls -l
total 0
cr--r--r-- 1 root sys 64 0x010000 Nov 25 2002 group
br--r----- 1 root sys 64 0x010001 Nov 25 2002 lvol1
brw------- 1 root sys 64 0x010001 May 3 15:42 lvol2
cr--r----- 1 root sys 64 0x010001 Nov 25 2002 rlvol1
crw------- 1 root sys 64 0x010001 May 3 15:42 rlvol2

[Devender Khatana]

Hi,

Just do lvremove /dev/vg01/lvol1 then you should be able to make a lvol with this name. Allthough the volume name does not make any sense except general naming convention as ultimately you are concerned with mount point. This means you can have /dev/vg01/lvol1 can be mount as /data and /dev/vg01/data also can be mounted as /data. Your applications see only /data and they do not concern from where it is mounted.

HTH,
Devender

[Nguyen Anh Tien]

Can you post the result of following command
#lvlnboot -v
#bdf
#vgdisplay -v vg00
I will see the result and tell you what should do next.