new 36 gig disk

[Ragni Singh]

hello everyone,

I have just installed 8 new 36 gig disks into my fc-10 machine and I would like to configure it at the volume group level. Is there a procedure of how its done as far as the steps I may need to take at the vg level. Your time and help will be greatly appreciated.

[nancy rippey · Accepted solution]

To create the volume group, logical volume and vxfs filesystem, also creating a mirror.

Creating Volume Group and Logical Volumes

pvcreate /dev/rdsk/cXtXd0 - For each disk

mkdir /dev/vgname
mknod /dev/vgname/group c 64 0x010000 - vg00 is 00, additional is 01 and so on

vgcreate /dev/vgname/dev/dsk/cXtXd0 /dev/dsk/cXtXd0 ... - List drives to be included

To create new logical volumes

lvcreate -n lv## /dev/vgname - Create a zero sized logical volume,

lvextend -l <# 4MB extents> /dev/vgname/lv## /dev/dsk/cXtXd0 - Make lv's size you want

Add disks to existing Volume Group

vgextend /dev/vgname /dev/dsk/cXtXd0

Mirror existing Logical Volumes

lvextend -m 1 /dev/vgname /dev/dsk/cXtXd0 - where cXtXd0 is disk adding mirror to

Remove Mirroring

lvreduce -m 0 /dev/vgname/lv## /dev/dsk/cXtXd0 - where cXtXd0 is disk removing mirror from

To create and mount a new filesystem
mkdir /mntpoint

newfs -F vxfs -b /dev/applvg/lvol_

mount /dev/vgname/lvol_ /mntpoint

Add entry to /etc/fstab

/dev/vgname/lvol_ /mntpoint vxfs rw,suid,delaylog,detainlog 0 2

[Jason VanDerMark]

The only thing that I would add is probably very obvious and that would be to ensure that the system sees the disks by running 'ioscan -fnC disk' and then create device files for the devices with 'insf -e'. Hope this helps.

Regards,
Jason V.

[Ragni Singh]

Now my other question here is that, right now we have a 36 gig disk but there is a possibility that we will be going to a 72 gig disk in the near future. How do I create the volume group with the -s option so it can recognize the 72 gig disk. Thanks for all the help

[James R. Ferguson]

Hi:

It is imperative that you create the volume group with the understanding that the 'max_pe' ('-e' option) ranges from 1 to 65535 extents but defaults to 1016 or the *largest* number of physical extents on the initial disk.

Once set, the 'max_pe', 'max_lv', 'max_pv' and 'pe_size' cannot be changed.

These parameters define the geometry of the LVM structure and must define a structure that can fit into one physical extent.

Regards!

...JRF...

[Sachin Patel]

Hi Sanman,
We had 36Gb disks then we replace to 72Gb now we are replacing with 181gb disks array.

We generally create the new array(lv,vgs)on another system with same mount point but more space then transfer the data from old vgs to new vgs on new system. then shutdown the original system and hook up the new array.

Note:scsi id on controller needs to be same.

Sachin