Node cloning

[Trent Piepho]

On Fri, 6 Apr 2001, Felix Rauch wrote:
> On Thu, 5 Apr 2001, Robert G. Brown wrote:
> [Copying /dev/hda to /dev/hd?]
> > One of many possible problems, actually.  This approach to cloning
> > makes me shudder -- things like the devices in /dev generally have to
> > built, not copied, there are issues with the boot blocks and bad block
> > lists and the bad blocks themselves on both target and host.  raw
> > devices are dangerous things to use as if they were flatfiles.
> 
> Unfortunately I'm not an expert in disk technology, so I might be
> wrong here... but I thought that the bad block lists were maintained
> by the disks themselves and not transparent to the OS.

Depends on the device.  For old stuff like MFM and RLL drives from the 80s and
floppy disks, the bad blocks are maintained by the OS.  For new drives like
SCSI and IDE, the bad block list is maintained by the drive itsself.

One problem with cloning a block device is that you have to copy the whole
thing.  You can't buy cheap IDE disks under 20GB anymore, but the initial
install on your nodes is probably a fraction of that.  I just built some nodes
up with 29GB drives, less than 500MB are used.  It takes a lot longer to copy
29GB than to copy 500MB.

I just wrote a small script that would partition the disk, format the
partitions, mount them, untar a system image, run lilo, generate ssh keys, and
run vi on the files I needed to add the hostname to. 

I would plug a new hd into the 2nd IDE controller on my prep machine, run the
script, and in about 5 minutes the drive would be ready to install.