Help Failing Disk Problem

James jamesh at lanl.gov
Mon Nov 5 13:16:10 PST 2007


On Mon, 2007-11-05 at 14:04 -0700, Warren Block wrote:

> On Mon, 5 Nov 2007, James wrote:
> 
> > On Mon, 2007-11-05 at 11:53 -0800, Sean Murphy wrote:
> >
> >> I have a FreeBSD 6.2 Release box with a single ide that has user data
> >> and the FreeBSD OS on a hard disk that is failing.  I need advice on the
> >> best way to clone the entire disk (or at least the data) onto a larger
> >> ide disk drive, then pull the failing disk and replace it with the
> >> clone.  What is the best way in FreeBSD to do that?
> >
> > The best way is to do it regularly before the hard drive is failing.
> >
> > Given that you haven't done that, there're a few methods. I'm a big fan
> > of rsync,  which is the nectar of the gods, but a lot of folks seem to
> > prefer dd for this kind of thing.
> 
> rsync is too high-level, and may not do exactly the right thing with 
> links or sparse files or who knows what. 

rsync -cav takes cares of symlinks and all that just right. It's a
beautiful thing.

Checksumming, too. Ah, bliss.


>  dd is too low-level--you get 
> the same partition table/bsdlabel and the exact same slice/partition 
> sizes.  That's okay on an identical hard drive, but a pain on one that's 
> larger.
> dump, on the other hand, is just right.
> 
> -Warren Block * Rapid City, South Dakota USA


dump has the problem that a lot of tools have, though, including rsync.
It creates a file list to start from. 

If the file names on the drive change during the dump, corruption can
occur. At least on linux. I remember Torvalds ranting about it on a
mailing list. I imagine FreeBSD suffers the same issue, though, as it's
a pretty generic problem.

dump is a good tool, though, no arguments really here.

James



More information about the freebsd-questions mailing list