Date: Tue, 29 Oct 2002 14:39:24 -0800 (PST) From: Matthew Dillon <dillon@apollo.backplane.com> To: "Jack L. Stone" <jackstone@sage-one.net> Cc: Mike Hogsett <hogsett@csl.sri.com>, "Mr. Darren" <darren780@yahoo.com>, freebsd <freebsd-stable@FreeBSD.ORG> Subject: Re: copy harddrive image Message-ID: <200210292239.g9TMdO64011366@apollo.backplane.com> References: <200210292048.g9TKmG1J015723@axp.csl.sri.com> <3.0.5.32.20021029162139.010ec928@mail.sage-one.net>
next in thread | previous in thread | raw e-mail | index | archive | help
:Hi, Matt: I've never tried your "cpdup" program I just installed the port
:to try it.
:....and wonder:
:
:1) why you prefer it to ole dump/restore?
cpdup is not a replacement for dump/restore. That is, the programs do
different things. cpdup does not produce an archive, for example,
it's just a glorified copy that is able to retain softlinks, hardlinks,
and other special features of files. On the otherhand, cpdup works
quite well on live filesystems and cpdup works incrementally, meaning
you can ^C it and restart it at any time and it will pick up where it
left off. dump does not work well at all on a live filesystem and
if you kill it you have to rerun it from scratch.
:2) what would be the typical syntax of the line command you use to copy
:over the ad0s1f to ad1s1f partition for example.
:
:I like your dd/cpdup combo approach and would like to try it on a test box.
:
:Best regards,
:Jack L. Stone,
:Administrator
Just do a 'man cpdup', or run cpdup without options. I typically
just use 'cpdup srcdir destdir'. Remember, cpdup will remove items
from the destination that do not exist in the source. It will ask
first by default.
Fdisk/disklabel/newfs the new disk
Mount newdisk partitions in /mnt, mimicking your normal mounts. for
example:
mount /dev/ad1s1a /mnt
mkdir /mnt/usr
mount /dev/ad1s1d /mnt/usr
mkdir /mnt/var
etc...
cpdup / /mnt
cpdup /usr /mnt/usr
cpdup /var /mnt/var
(go into single user)
cpdup them all again
shutdown, swap disks, reboot.
For dd the person who posted the comment in regards to block size is
absolutely correct. It is safest to run dd only on idle
filesystems, leaving the source filesystems mounted and the destination
partitions unmounted (of course). This way if you accidently specify
a source filesystem as a destination the OS will not allow the source
to be openned for writing (it will return a 'device busy' error),
which can save your ass if you make a mistake. But it also means you
have to be sure the source filesystem are idle and sync'd up
(sync; sync; sync; sync; ...), and that you have to fsck the
destination partitions after the dd finishes.
# example
dd if=/dev/ad0s1a of=/dev/ad1s1a bs=1m
Using dd can be dangerous. Be sure not to accidently specify a
destination that winds up overwriting your source drive! Leaving
the source mounted mitigates the risk of accidently blowing something
important away.
-Matt
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200210292239.g9TMdO64011366>