Disk Enclosures
1748275 Members
3768 Online
108761 Solutions
New Discussion юеВ

Storage migration..

 
SOLVED
Go to solution
Shankar_6
Regular Advisor

Storage migration..

Hi All,

Currently we have 2*node cluster running OPS.
Disks are accessed from XP-512.

We have a plan to migrate the stroage to Hitachi.

I just would like to know action plan if anything available for this storage migration..(like moving data to new storage,LVM configurations,cluster lock config change..etc..).I too know it is hard to get info as its not a simple task.

I am asking this to get an idea for the change.

shankar.
6 REPLIES 6
freddy_21
Respected Contributor

Re: Storage migration..

what is your method? is it mirroring or hitachi tools or any else?
Shankar_6
Regular Advisor

Re: Storage migration..

Hi,

We have a plan to use method of LV mirroring.

waiting for reply..

Shankar
freddy_21
Respected Contributor
Solution

Re: Storage migration..

My assumption your cluster have one package.


At server one:

1. Make sure about size of your new disk. it has same size with your current disk?
2. pvcreate your new disk ( hitachi disk )
pvcreate -f /dev/rdsk/cxtydz
3. extend new disk to vg configuration
vgextend /dev/vg01 /dev/dsk/cxtydz
4. create lvmpvg file
lvmpvg will be help you to mirror to new disk.
ex:
VG /dev/vg01
PVG PVG0
/dev/dsk/c4t0d0
PVG PVG1
/dev/dsk/c5t0d0


PVG PVG0 --> contain all disk XP at vg01
PVG PVG1 --> contain all disk hitachi at vg01


5. run lvextend command to do mirror

lvextend -m 1 /dev/vg02/lvol1


6. After sync, shutdown your package and your cluster. Activate volume group at node oneand remove your xpdisk
vgchange -c n vg01
vgchange -a y vg01


lvreduce -m 0 /dev/vg01/lvol1 /dev/dsk/cxtydz

after all disk xp already unused for logical volume, remove disk xp
vgreduce /dev/vg01 /dek/dsk/cxtydz

7. export and import volume group.

vgexport -v -s -m vg01.map vg01
copy to another node

at server 2:
remove all volume group at the cluster configuration
vgexport vg01

import map file

mkdir /dev/vg01
mknod /dev/vg01/group c 64 0x010000
vgimport -v -s -m vg01.map vg01

now both your node have same lvm configuration


At server one:
8. changes your cluster ascii file for device cluster lock disk

9. apply ascii file that you already edited

10. deactivate vg
vgchange -a n vg01

11. run cluster
cmruncl


Thanks
Freddy


Shankar_6
Regular Advisor

Re: Storage migration..

Hi,

Many thanks for ur detailed steps.

Do we need to create lvmpvg file.?right now the lv's are not mirrored..and allocation policy is set to "Strict".

Please clarify..

Regards
Shankar
Chan 007
Honored Contributor

Re: Storage migration..

Shankar,

Is you are going to have your mirrors always it is fine to create a LVMPVG file else if this is just an one off, leave it.

Also your profile is like "I have assigned points to 23 of 110 responses to my questions. "

This forum needs just appreciation by providing points. But your provide is very low. See fredy's responce, he would have spent more than 10 to 15 Mins for you. Just as a courtesy provide points to any one who helps you.

This will helps this forum.

Chan
freddy_21
Respected Contributor

Re: Storage migration..

Shankar,

For me, Point sometime really meaningfull. It will be measurement for me about what i explain to you. You can understand or not. or you satisfied with my guidence or not.

how big you give point. we will appreciate.


Thanks
Freddy