Subject: Re: Odd ccd(4) problem
To: Bill Sommerfeld <sommerfeld@orchard.medford.ma.us>
From: Jason Thorpe <thorpej@nas.nasa.gov>
List: current-users
Date: 09/06/1995 08:54:19
On Wed, 06 Sep 1995 07:45:12 -0400 
 Bill Sommerfeld <sommerfeld@orchard.medford.ma.us> wrote:

 > Here's one idea...
 > 
 > All components of the ccd must have the same fstype; the fstype of the
 > ccd is that type.
 >
 > If the components don't match, you get an error when assembling the
 > ccd out of its component pieces.

That seems reasonable, although there are still some fstypes one might want 
to exclude...

 > optional:
 > 
 > If you update the disklabel of the ccd with a different fstype, the
 > disklabels of the component pieces get updated.
 > 
 > i.e., if I newlfs a ccd formerly containing an FFS partition, the
 > component pieces should now all be relabeled FS_BSDLFS instead of
 > FS_BSDFFS.

There are pratical problems with this ... for one, when updating the 
fstype of a ccd partition, you'd have to calculate which components the 
partition lives on, and update only those.  What if a ccd partition 
occupies only one block of a component?  Yuck.  Also, there's the problem 
of ports that Just Can't change fstypes ...

--------------------------------------------------------------------------
Jason R. Thorpe                                       thorpej@nas.nasa.gov
NASA Ames Research Center                               Home: 408.866.1912
NAS: M/S 258-6                                          Work: 415.604.0935
Moffett Field, CA 94035                                Pager: 415.428.6939