Mike Mangino writes:
> I seem to remember hearing some talk about a Copy On Write filesystem
> that allowed cloning of files and copying the data on write to the file.
> Does anyone know where that work is? I've been thinking about adding
> support for this to ext2 so that it could be used in a development
> environment. It makes sense to me that development using CVS could save
> a lot of disk space if this were implemented.
We have developed the COW functionality as part of OBDFS (Object Based
Devices), and the Lustre clustering project. You can find out more about
it at our web site http://www.lustre.org/ but you should realise that
this is NOT production quality code. While it is based on ext2, if you
use the COW (snapshot) function, you can no longer fsck the filesystem
using regular e2fsck (well, you could but everything would blow up ;-).
There is also COW ability in LVM, but I'm not sure if it's ready for
production either, as there were some strange problems reported on linux-lvm
mailing list.
By all means, we would be happy if you had a look at OBDFS and snapshots,
but I wouldn't go putting your development repository on it.
Cheers, Andreas
--
Andreas Dilger \ "If a man ate a pound of pasta and a pound of antipasto,
\ would they cancel out, leaving him still hungry?"
-- Dogbert