Cloning boot drive - more details

nthwaver at nthwaver at
Tue Apr 25 22:41:57 UTC 2006

My system partitions (/, swap, /usr, /var, /home) are currently spread
onto a 10GB and a 20GB IDE drive, but I'd like to save space by
consolidating these along with some (not heavily accessed) data
partitions into a larger 250GB disk.  The other drives (at this point
a couple of SATA in RAID0) should be unaffected.  I'm a relative
newbie and although I've read the handbook and the past months'
threads regarding cloning, I still have a few questions.

1) Am I correct in understanding that I can simply connect the new
drive to a spare IDE controller and boot from the old disk, using
sysinstall to make the new partitions and give them temporary mount
points (choosing "yes" to install bootmanager), then dump | restore to
move each FS, and simply take out the old drives and switch over to
the new one?  Will this boot and run seamlessly?  At what point should
I edit the old /etc/fstab that was copied over?  If I *can* do this,
then what are the benefits of doing a fresh install on the new drive

2) If I dump | restore from a *running* system, will the resulting
clone be confused when it's booted up?  Are any crucial changes or
balancing acts made upon shutdown that the new drive will miss?  Or,
is its main purpose fulfilled when it's loaded into memory on boot?

3) The handbook also recommends using boot0config, but how necessary
is this if I just plan on simply replacing the original drive?

4) How are the prospects of data recovery affected by FreeBSD's use of
"slices" for filesystems on top of partitions?  Experience tells me
that with traditional partitions, a corrupted file tree or data in one
area needn't prevent retreival of the other areas.  Is this so with
"slices" as well?

Thank you very much,

More information about the freebsd-questions mailing list