Reconstructing RAID-0 to RAID-10 using omconfig
Harald_Jensas at Dell.com
Harald_Jensas at Dell.com
Sat Jun 9 06:55:26 CDT 2007
> -----Original Message-----
> From: linux-poweredge-bounces at dell.com [mailto:linux-poweredge-
> bounces at dell.com] On Behalf Of John Buttery
> Sent: 08 June 2007 18:56
> To: linux-poweredge-Lists
> Subject: Re: Reconstructing RAID-0 to RAID-10 using omconfig
> * On Friday 08 June 2007 10:47, Tino Schwarze
> <linux-poweredge.lists at tisc.de> wrote:
> >Then use dd to copy the whole filesystem as is. If your network is
> > secure, netcat is a nice tool (sometimes just called nc).
> > nc -l -p 4242 > fsdump.gz
> >on the backup server to start listening on port 4242.
> >Then run
> > dd if=/dev/your/device of=- bs=1M | gzip -1 | nc $backupserver 4242
> >to transfer everything gziped (depending on network and CPU speed,
> > gzip may or may not speed things up - try dd with count=100 to look
> > what it takes to transfer 100MB).
> >Later, do it the other way around: Run
> > nc -l -p 4242 | gunzip | dd if=- of=/dev/new/device bs=1M
> >on the server, use
> > cat fsdump.gz | nc $server 4242
> >to transfer the dump via network.
> First off, thanks for the info.
> 1) Will dd work when the partition isn't the same size? The process
> here would be to dd the filesystem out, re-initialize the array (at
> which point it will double in size), then dd the filesystem back in.
> Is dd going to be OK with that?
> 2) Do I run this once per partition, or once for the whole disk? In
> other words, is it gonna be "dd if=/dev/sda1 ..." followed by "dd
> if=/dev/sda2 ...", etc ... or just one big "dd if=/dev/sda ..."?
> 3) I don't even need the network -- the temporary storage location is a
> USB disk connected directly to the server in question -- but I'll keep
> that in mind for next time. :)
> John Buttery <johnb at academicsuperstore.com>
> System Administrator
> Linux-PowerEdge mailing list
> Linux-PowerEdge at dell.com
> Please read the FAQ at http://lists.us.dell.com/faq
I would sooo opt for the command you suggested yourself in the beginning, mind changing the size parameter:
omconfig storage vdisk action=reconfigure controller=0 vdisk=0 raid=r10 size=264g adisk=0:0,0:1,0:2,0:3
The size parameter is the total size of your virtual disk. In the case of a RAID 10 it would be:
(("Physical Disk Size" * "Number of Disks) / 2) = "Virtual Disk Max Size"
In your case close to 264 GB right?
The command will not wipe any data of the drives.
This will, change the RAID Volume from your RADI 0 to a RAID 10 live. No downtime, except a reboot once it is done to get Linux to recognize the new drive size once you are done. In windows a simple rescan in Disk Management would suffice, but I do not think that can be done in Linux? At least not for a device with a mounted filesystem. Thus the reboot. Of course if you do not have OMSA installed on the system you need to install it or boot to a LiveCD with omsa to perform this operation.
PS! This operation will take a long time to complete. But it is defenetly easier than imaging to a different drive with dd and then reimage after...
Also, dependinig on the version of OMSA you use the adisk parameter might be pdisk. I belive it has been pdisk since 1 or 2 releases back.
More information about the Linux-PowerEdge