On 13-Feb-00 Larry McVoy wrote:
> The drive venders will obviously go
> through a progression to get there (notice that Quantum bought Meridian?).
> If the market will support a $1K drive + CPU + NIC, then putting the CPU +
> NIC down on the drive directly is just a standard cost reduction.
I think this makes a huge amount of sense for drive vendors, because it means
they can sell drives to people who don't like opening their machines.
On the other hand, I don't think it makes any sense to buy them.
For home use, maybe, but I can't think of anything worse to administer than a
network with hundreds of piddling little 20GB disks all over the place. People
learned that having lots of diskful workstations was a complete pain to
administer; having lots of anonymous network disks scattered in cupboards all
over the place is even worse.
I agree that things like fibrechannel are a waste of time and money, and
there's no reason to not use 100/1000Mbit ether for disk attachment (if there's
problems, fix them; don't invent some new expensive technology). Similarly,
STP looks like a very good way to talk to drives. However the way the drive
companies are doing it now (with their heads full of "maybe we can get more
than 1% margin!") is just crazy.
In other words, something like STP is exactly what they want, along with some
way of aggregating all these disks into a single logical unit as far as storage
and administration goes. Which means diskless (stateless, interchangable,
robust) Linux boxes doing NFS/SMB <-> STP and as many networked drives on the
back side as you want.
J
-
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/
This archive was generated by hypermail 2b29 : Tue Feb 15 2000 - 21:00:25 EST