On Thu, 2006-05-04 at 20:14 +0800, Wu Fengguang wrote:
> On Thu, May 04, 2006 at 09:33:24AM +0200, Arjan van de Ven wrote:
> >
> >
> > > Ascending block numbers on disk can be read very fast, as the disk needs no or
> > > less seeking. That's even true for stripes and mirrors. (I grant you that
> > > there are complicated setups out there, but these could be handled similar.)
> > >
> >
> >
> > btw this all really spells out that you may want to do this as a device
> > mapper thing; eg have a device mapper module that can do "lookaside" to
> > a different order/mirror block whatever. The filesystem just doesn't
> > want to know; do it at the DM level ;) That also solves the entire
> > caching layering problem etc ;)
>
> I guess some big corps might want to install such a layer into their
> storage products ;)
maybe, could be.
Doing it at this level also has the advantage that fs metadata becomes
seek free as well; since that is mostly fixed-location inside the fs,
reordering that on an fs-level is really hard.. on a dm level.. easy.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
[Index of Archives]
[Kernel Newbies]
[Netfilter]
[Bugtraq]
[Photo]
[Stuff]
[Gimp]
[Yosemite News]
[MIPS Linux]
[ARM Linux]
[Linux Security]
[Linux RAID]
[Video 4 Linux]
[Linux for the blind]
[Linux Resources]
