Disklabe oddity

[Guy Van Sanden]

On Sun, 2003-09-21 at 02:46, Greg 'groggy' Lehey wrote:
> [Format recovered--see http://www.lemis.com/email/email-format.html]
> 
> Still wrapped output.  This is painful to recover, and I tend to lose
> interest when it continues.
> 

I looked at your page, and you make a good point.
When I return later today, I'll look/google for a way to set Ximian
Evolution up to do this correctly (if that is possible), otherwise, I'll
file a bug-report with them.

> On Saturday, 20 September 2003 at 11:53:41 +0200, Guy Van Sanden wrote:
> > Thanks you for your very complete answer Greg
> >
> >
> > On Sat, 2003-09-20 at 10:59, Greg 'groggy' Lehey wrote:
> >> [Format recovered--see http://www.lemis.com/email/email-format.html]
> >>> The disk was formatted by the FreeBSD install procedure.
> >>
> >> If you haven't changed anything here, it would be interesting to know
> >> in more detail just what you did.  To judge by the surprising number
> >> of partitions, you didn't take the defaults.
> >
> > I just created seperate partitions for / /tmp /var /usr etc.  The c
> > partition was created by FreeBSD on its own.  Could this be a BIOS
> > problem, my system BIOS predates that size of disks by far?
> 
> Barely possible.  
> 
> >>> The system is an older Digital PC (3500) PII 333 Mhz.  The disk is
> >>> a 40 GB IDE drive (WD)
> >>>
> >>> BTW, I'm looking for a safe way to 'grow' my rootfs, I've looked
> >>> arround before, but I'm still not clear on the right procedure for
> >>> it.
> >>
> >> Take a look at growfs(8).  To do it right, you need space directly
> >> behind the root file system.  Even Vinum won't help here.  You could
> >> move the swap space elsewhere, for example.
> >
> > I'm looking at that option, lucky that I have a second disk with rsynced
> > mirrors of all partitions on the first one.  I can just remove /home and
> > /data and copy them back later.
> 
> Well, yes, or you can completely reinstall.  I was looking at a less
> intrusive way of doing it.
> 
> > Perhaps I can add the swap space to / and create a new swap further
> > back on the disk.
> 
> It's not a good idea to add swap space to a file system.  It's better
> to have your own partition.  Probably what you have is more than
> adequate, though.

I wasn't planning to put the swap on the fs, I would create a swap
partition further back on the disk, then delete the existing one and
extend the root fs to include the that space.
I'm only not sure if this would cause the slice letters to move (/ is
slice a, swap is b - it I remove it, will it cascade the others down?)

> 
> >> You can clear the error by running disklabel -e /dev/ad0s1a in single
> >> user mode, and changing the length and offset of partition c (offset
> >> 0, add 63 to the size).
> >
> > I wanted to try this on the mirror disk first, it also shows the offset
> > at 63 using disklabel -r
> > Yet doing disklabel -e on it shows the offset at 0
> > 8 partitions:
> > #        size   offset    fstype   [fsize bsize bps/cpg]
> >   a:   327680        0    4.2BSD     2048 16384 20488   # (Cyl.    0 - 325*)
> >   b:  1007984   327680      swap                        # (Cyl.  325*- 1325*)
> >   c: 80418177        0    unused        0     0         # (Cyl.    0 - 79779*)
> >   d:  1007616  1335664    4.2BSD     2048 16384 62984   # (Cyl. 1325*- 2324*)
> >   e:  1024000  2343280    4.2BSD     2048 16384 64008   # (Cyl. 2324*- 3340*)
> >   f: 18120704  3367280    4.2BSD     2048 16384 28552   # (Cyl. 3340*- 21317*)
> >   g: 20971520 21487984    4.2BSD     2048 16384 28552   # (Cyl. 21317*- 42122*)
> >   h: 37958673 42459504    4.2BSD     2048 16384 28552   # (Cyl. 42122*- 79779*)
> 
> It looks as if you have two different partition tables.  This one
> doesn't match the other.  Are you still getting the message?

Strangely engough, yes
Someone suggested that this might be a problem in -current, so I'll
check those lists later today.

> 
> Greg
> --
> When replying to this message, please copy the original recipients.
> If you don't, I may ignore the reply or reply to the original recipients.
> For more information, see http://www.lemis.com/questions.html
> See complete headers for address and phone numbers