HEADSUP: geom_vinum committed

[Lukas Ertl]

On Tue, 15 Jun 2004, Paul Mather wrote:

> # Commented lines are from disklabel, and used to position vinum
> # subdisks.  Note that the driveoffset is 16 sectors less than
> # the offset in the disklabel, to account for the bootstrap
> # (to be in accordance with the start of the vinum partition).

Does that mean that you're mixing bsdlabels and vinum?  This could 
certainly lead to problems.  Furthermore I haven't tested root-on-vinum 
yet.

> (BTW, gvinum failed to parse the above config file correctly.  I got
> complaints about "setupstate" from what I rememeber.  Vinum accepts it
> without complaint.)

That's because I haven't implemented the 'setupstate' keyword.  In my 
opinion any RAID that needs to be initialized should be initialized.  We 
could argue about that, of course. :-)

> (Note that the "driveoffset" values for the subdisks matches those
> given in the configuration file.)

I'll try to replay that, but it could take some time to setup a 
root-on-vinum configuration here.

> Needless to say, those testing out gvinum on existing Vinum
> configurations should exercise due caution due to the real possibility
> of corruption (as Lukas warned in his initial announcement).

That's what I said, yes.

cheers,
le

-- 
Lukas Ertl                         http://homepage.univie.ac.at/l.ertl/
le at FreeBSD.org                     http://people.freebsd.org/~le/