Destroy GPT partition scheme absolutely, how?

[John Baldwin]

On Tuesday, October 04, 2016 06:23:26 AM Warren Block wrote:
> On Mon, 26 Sep 2016, John Baldwin wrote:
> 
> > On Tuesday, September 27, 2016 12:36:22 AM Ngie Cooper wrote:
> >>
> >>> On Sep 26, 2016, at 22:48, Ernie Luzar <luzar722 at gmail.com> wrote:
> >>
> >> ...
> >>
> >>> This little script has been posted before. Maybe it will be what your looking for. Called gpart.nuke
> >>>
> >>> #! /bin/sh
> >>> echo "What disk do you want"
> >>> echo "to wipe? For example - da1 :"
> >>> read disk
> >>> echo "OK, in 10 seconds I will destroy all data on $disk!"
> >>> echo "Press CTRL+C to abort!"
> >>> sleep 10
> >>> diskinfo ${disk} | while read disk sectorsize size sectors other
> >>> do
> >>> # Delete MBR and partition table.
> >>> dd if=/dev/zero of=/dev/${disk} bs=${sectorsize} count=1
> >>> # Delete GEOM metadata.
> >>> dd if=/dev/zero of=/dev/${disk} bs=${sectorsize} oseek=`expr $sectors - 2` count=2
> >>> done
> >>
> >> Why not just use "gpart destroy -F provider"?
> >
> > That doesn't always work.  In particular, if a disk was partitioned with GPT
> > and then you use normal MBR on it afterwards, the 'gpart destroy -F' of the
> > MBR will leave most of the GPT intact and the disk will come up with the old
> > GPT partitions, not as a raw disk.
> 
> Right.  So do a gpart destroy -F of whatever is on there, ignoring 
> errors, then a gpart create -s gpt.  Now there is definitely a secondary 
> GPT, and a final gpart destroy -F removes it cleanly.

It is usually simpler to just dd zeroes over the first N and last N sectors.
The only fool-proof way to ensure you don't have dangling tables in the
middle of the disk is to zero the entire disk, but that takes too long.

-- 
John Baldwin