Not too familiar with VxVM. We have a Device c1t0d0, disk rootdg01, group rootdg that was paired with Device c1t1d0, disk mirror_1, group rootdg.

c1t0d0 has failed.

When we get a replacement for c1t0d0 can we just run vxdiskadm and choose choice #5 to "replace a failed or removed disk" ?

Will it be that simple?

Thanks!

Marty

Not quite. You run that AFTER replacing the failed disk to start using the new one. BEFORE you need to choose the "remove a failed disk for replacement" option (or something very similar to that). You need to run vxdiskadm before the physical hardware replacement and again after.

Replacing a mirrored root disk under VxVM is a little different from replacing any other mirrored disk under volume manager
First of all you have to confirm that your system cn boot from the mirrored disk, To ensure that run the following command out from my system is appended below (truncted for clarity)


root@baluqorcz1:/# eeprom
nvramrc=devalias vx-rootdg01 /pci@9,600000/SUNW,qlc@2/fp@0,0/disk@w2100000c50850e70,0:a
devalias vx-rootdg02 /pci@9,600000/SUNW,qlc@2/fp@0,0/disk@

You should create an alias to boot from alternate boot disk.
Once that is done, you have to remove the disk from volume manager control example
$vxdisk offline c1t0d0d0
then to physical pull the bad disk out you have remove it from OS control example
#luxadm removez-device c1t0d0s0
then you are safe to remove the disk
after that run#devfsadm -C -c disk
so that the system rereads all the drives present
then install the new drive.
After that make the OS presence of a new drive for that run this command
#luxadm insert_device c1t0d0s0
then again run this command #devfsadm -C -s disk
then bring the disk under volum manager conrol example
#vxdisk online c1t0d0s0
check that the volume manager can see it by running
#vxdisk list
now run #vxdiskadm command

You also might consider reading Veritas Volume Mnager Trouble shooting guide section Re-Adding and replacing Boot Disks

Thanks very much to both of you for taking the time to respond to my query. I'll try to post details of this activity after I get a new replacement disk. Wish I had a place to experiment . We're having a little trouble getting the replacement disk on-site so my feedback may take a few days.

I will try to locate and read The section in "Veritas Volume Mnager Trouble shooting guide" (section Re-Adding and replacing Boot Disks). Thanks again!

Volume Manager Support Operations
Menu: VolumeManager/Disk

1 Add or initialize one or more disks
2 Encapsulate one or more disks
3 Remove a disk
4 Remove a disk for replacement
5 Replace a failed or removed disk
6 Mirror volumes on a disk
7 Move volumes from a disk
8 Enable access to (import) a disk group
9 Remove access to (deport) a disk group
10 Enable (online) a disk device
11 Disable (offline) a disk device
12 Mark a disk as a spare for a disk group
13 Turn off the spare flag on a disk
14 Unrelocate subdisks back to a disk
15 Exclude a disk from hot-relocation use
16 Make a disk available for hot-relocation use
17 Prevent multipathing/Suppress devices from VxVM's view
18 Allow multipathing/Unsuppress devices from VxVM's view
19 List currently suppressed/non-multipathed devices
list List disk information


? Display help about menu
?? Display help about the menuing system
q Exit from menus

Select an operation to perform:

First do Option 4 (Remove a disk for replacement)

Then remove the bad disk/add new disk (luxadm remove_device /devfsadm etc ...)

Then do Option 5 (Replace a failed or removed disk)

Still waiting on the disk.

Isn't it necessary to run

vxdctl enable

once the operating system can "see" the new disk via the format command?

Also, does the fact that I already ran choice #11 (disable -offline- a disk device) change the instructions you provided?

Thanks!

I think you cannot do option 11 first.

The right way is to do opt 4, remove/add disk then opt 5.

if I can recall correctly when you do opt11 then you later try to do opt 4/5 it may fail or give some error. I think you need to try to bring the disk back online before doing option 4.

But if you cannot do opt4/5 there is a workaround from sunsolve.

Solution 212645 : How to replace a root drive in Volume manager if vxdiskadm option4/option5 is not working

Btw I've never attempted this procedure before, so use at your own risk.

======================
Removing the Root disk
======================

1) # vxdg -g rootdg -k rmdisk

2) # vxprint -htg rootdg
remove the plexs associated with the original disk (disabled nodevice)

3) # vxplex dis opt-01 (run against all plexes associated with root disk)
- this is a standard plex that is creatd in encapsulation

4) # vxedit -r rm opt-01 (run against all plexes associated with root disk)

5) # vxedit rm

6) # vxdisk list
should no longer be listed as a removed entry

7) # vxdisk rm c#t#d#s2 (root disk)



=======================
Replacing the root disk
=======================

1) preform drive replacement steps per system requirements
if drive has already been replaced, skip this step

2) # format
Delete all partitions except for slice 2, and label the disk

3) # vxdctl enable

4) # vxdisksetup -i c#t#d# (rootdisk)

5) # vxdg -g rootdg adddisk rootdisk=c#t#d#

6) # vxdiskadm - opt6 to mirror
choose root mirror as the disk to mirror
root disk is destination disk
this should recognize that it should be a root disk and write a
bootblock onto it

7) # prtvtoc /dev/rdsk/c#t#d#s2 (rootdisk)
check to make sure that there are regular partitions


Product

VERITAS Volume Manager 3.2 for Solaris 9
VERITAS Volume Manager 4.0 Software
VERITAS Volume Manager 3.5 Software
VERITAS Volume Manager 3.2 Software
VERITAS Volume Manager 3.1.1 Software
VERITAS Volume Manager 3.1 Software
VERITAS Volume Manager 3.0.6 Software
VERITAS Volume Manager 3.0.4 Software
VERITAS Volume Manager 3.0.3 Software
VERITAS Volume Manager 3.0.2 Software



Keywords

volume manager, rootdisk replacement fails



Previously Published As

75082



Attachments

This solution has no attachment

Well, it turns out that the system needed to be patched.

After many unsuccessful attempts to add the disk to the mirror, Sun Support discovered that we were experiencing a problem with VxVM v4.1 patch level 171080-04 which could be fixed with patch 171080-07. The patch was installed and option 5 of vxdiskadm allowed the new disk to be added and re-synced.

Thanks for your assistance!

Here is what I would do (CLI - manual method):
Remove the old vx objects totally from the cfg (clean it out):
If disabled (vxedit –g rootdg –r rm rootdg01-0x)
if active (vxplex –o rm dis rootdg01-0x)
same as above for remaining plexes (see vxprint -thr & vxdisk list output)
vxdg –g rootdg rmdisk xxxxxxxxx
/etc/vx/bin/vxdiskunsetup –C cxtxdx
vxdisk rm cxtxdx

Run format - zero all partitions except slice 2
Label the new disk under format
run vxdctl enable

Find privite region lenght on existing disk (to match)
run vxprint -thr
TY NAME ASSOC KSTATE LENGTH
sd rootxxxxxxx - ENABLED 2743
(alt command) vxprint -thr | grep sliced
DM NAME DEVICE TYPE PRIVLEN PUBLEN STATE
dm rootmirror c1t1d0s2 sliced 2743 71681511 -

/etc/vx/bin/vxdisksetup –i cxtxdx puboffset=2744 (“rootdiskPriv” length + 1)
Note: Substitite the option old_layout for puboffset above for patch level 1 and above. (output: modinfo | grep vx = 3.2t_p1+)
So the command above would be (without puboffset statement):
vxdisksetup -i cxtxdx old_layout
vxdg –g rootdg adddisk rootdisk=cxtxdx
/etc/vx/bin/vxrootmir rootdisk
vxassist mirror swapvol rootdisk
vxassist mirror opt rootdisk
vxassist mirror usr rootdisk

I know that this works as I have used it many times in my env.
Please don't take my word for this - you need to test in your env.
Test on a test system by pulling a mirrored disk on a test box.
good luck