[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

The Fine Art of Making a Bootable Drive; more



I am the one who posted stating that I can't seem to make a
bootable new hard drive for my Linux Squeeze system. It's been
quoted, "It ain't what you don't know that will hurt you, but
what you know that just ain't so." I think I am in that
territory now. What I have been doing was to format the new
active partition (sda1) with ext4 or ext3, using the rest of the
drive space as extended primary Partition 2 and overwriting with
a logical Partition 5 for swap.
	I then would use rsync to copy all of the old drive
including special files to the new drive and one could see /dev
and all hard links to initrd.gz where they should be. The final
step which seems to be the kiss of death is to use grub-install
on the rescue disk to put a MBR on /dev/sda.
	I never got this to work. Today, I did

dd if=olddrive of=newdrive which worked but produces a logical
drive exactly the same size as the old drive which is 10 GB.
Since the new drive is 16 GB, that wastes 6 GB of capacity which
is why that is not ultimately acceptable. Just for fun, though,
I tried the truncated new drive and the system booted right up
which proves that it is not hardware. It is the way I am
building the drive.
	I can probably use tune2fs to re-size this new drive by
blowing away the extended swap partition, moving the upper
boundary of Partition 1 to 15 GB and then making a smaller
extended Partition 2 with swap overwrite, but I am curious as to
why the first method simply has never booted?
	The bad drives I created for Debian Squeeze were also
formatted with ext4 for one attempt and ext2 for a different
attempt. This is because the new drive is a flash drive and will
need to be mounted to take wear and tear of multiple writes in
to consideration. The old drive was formatted ext3 so the
truncated new drive is presently also formatted the same way.
	Now that I know the hardware is not causing the issue, I
just need to understand why or at least get some theories as to
why what seems like a logical way to build a drive doesn't work.

Martin McCormick


Reply to: