Re: New Solaris 7 features

Malcolm Beattie (mbeattie@sable.ox.ac.uk)
Wed, 11 Nov 1998 10:36:02 +0000 (GMT)


MOLNAR Ingo writes:
>
> On Tue, 10 Nov 1998, Malcolm Beattie wrote:
>
> > While I'm here: an idea for providing a snapshot capability to any
> > block device (allowing you to snapshot then backup any fs you like).
[...]
> > snap keeps a dirty bitmap and a sparse block structure on the spare
> > block device. Writes to snap mark the block in the dirty bitmap and
> > store the original block in the sparse structure. Reads from snap
> > check the dirty bitmap and read from the original or the sparse block
> > map.
>
> we do not want to write to the 'snap'. We want to use the 'snapshot
> device' to carry a snapshot for some limited (or unlimited) time, eg. for
> backup purposes. This thing is completely transparent to the 'normal'
> device.

Slip of the keyboard, sorry: I meant writes to the snapshotted device
(not snap itself) are intercepted and written to the spare device,
same as you describe. Good to know your md will be doing that. Could
you elaborate on whether the spare device that you're writing to will
be a plain log of all original blocks, a sparse map like I was talking
about or something entirely different? It would be nice to be able to
snapshot large devices while only using a smallish spare device that
is just large enough to hold the deltas while the backup or suchlike
is taking place.

--Malcolm

-- 
Malcolm Beattie <mbeattie@sable.ox.ac.uk>
Unix Systems Programmer
Oxford University Computing Services

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/