Re: strategy for disk replacement?
- To: debian-powerpc list <email@example.com>
- Subject: Re: strategy for disk replacement?
- From: "Matthew T. Atkinson" <firstname.lastname@example.org>
- Date: Thu, 19 Aug 2004 17:35:05 +0100
- Message-id: <1092933305.22219.154.camel@localhost>
- In-reply-to: <1092920454.4062.16.camel@pismo>
- References: <1092920454.4062.16.camel@pismo>
On Thu, 2004-08-19 at 14:00, david howe wrote:
> Can I rsync the disk to a remote and then copy back onto the new disk?
> Does anyone have any better suggestions?
Yes, I recently used this technique to recover a dead server because it
died in the middle of a dd (image) backup :-).
You'll need options such as:
--verbose --recursive --perms --owner --group --stats --progress
--delete --exclude=/proc --one-file-system --sparse --times --devices
--links --hard-links --delete-excluded
Remember to remake the proc directory in the tree you copy it to (sys as
well if you're using 2.6).
If you tar up this backup, make sure that you use the correct tar
options (man tar helps).
Regarding imaging, AFAIK it is very hard to resize partitions
afterwards, so you'd be wasting 30GB of space (unless you used fdisk to
create a partition in the free space) if you used dd. I am sure there
are more capable image programs but this rsync technique does work if
If you did want to use dd, you could probably use it in conjuction iwth
netcat and a FIFO to send out the resultant image over the network to
another computer for archival. But I've never attempted such a thing
Hope this helps, bye just now,
Matthew T. Atkinson <email@example.com>