Hi, been asked a question on upgrading 4 x 146Gb in a RAID 5 setup disks to 4 x 300Gb disks.
It is a standalone ESX host where ESX is installed to its own mirrored disks and VM's currently held on the 4 x 146Gb disks as mentioned above. They want to upgrade the VM disks to 300Gb disks.
Currently we plan to use FastSCP to transfer the VM data whilst switched off to an NTFS partition that it can talk to on the network. Replace the disks and configure them for use then transfer the VM data back. This obviously would require a fair period of downtime on the VM's.
I am wondering whether I am missing a trick a here and if it can be done quicker and better. Does anyone have any suggestions.
Does your Server not have spave to add the additional disks locally?
This will be quicker than copying accross the network.
If not, can you add the disks to a different server to migrate the data, then swap the disks out?
Your biggest problem is that you are currently proposing 2 sets of copies . . .
Going out on a limb . . I wonder whether you could swap out 1 disk at a time and let the RAID parition regenrate (I presume Raid 5 for the 4 disks), then when all 4 disks are 300 extend / add the additional space andf use an extent?
Hello Bullet, thanks for the reply.
Unfortunately there is no spare space as this would be the obvious way.
Not got another server to add them to.
Yeh the 2 copies is the reson for me looking for alternatives.
You read my mind as I thought about doing the swap out 1 at a time. My prob here is the setup is rubbish and no backups running and as I haven't actually done that before upping them in a current RAID 5 set I would want a full backup taken and that currently involves the first part which is to copy them off.
Appreciate your thoughts though
OK, without a backup . . I'd defo say the way to go is SCP copy the data, upgrade and SCP it back . . .
You'd be crazy to risk any in place updates without a backup.