--Bzq2cJcN05fcPrs+
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline

On Tue, Feb 15, 2005 at 04:21:12AM -0800, Gordon Waidhofer wrote:
> The original post was for ATA over *ethernet*, not IP.
> I looked at the spec. It is extremely light weight.
> Conceivably it could be implemented in an ASIC.

iSCSI has a lot of promise, both in terms of cheaper SAN
infrastructure and in terms of wide-area functionality.  It's
.. debatable .. how much of that promise has yet been realised, or
will be soon.

One argument in that debate might be that iSCSI implementation costs
and development times are too high for a majority of local SAN backend
storage interconnect, particularly in the high-capacity
performance-noncritical "near line" storage market for large image
archives, backups, etcetc which are already presenting arrays of ATA
disks as FCAL to SAN raid controllers.  It's turning out that special
'storage rated' ethernet switches are needed for iSCSI, or at least
vendors seem to be trying to convince customers so.  FC-AL is still
the local back-end SAN interconnect of choice, despite the promises of
iSCSI, and that comes with costs. On their sales quotes for the other
side, iSCSI replaces somewhat expensive FCAL HBA's with even more
expensive HBA's that offload iSCSI and TCP processing to the card.

If you want all the performance and throughput all the way through,
you pay that price.  If you don't need it, but you still want some of
the virtualisation flexibility and/or bulk capacity, options and
offerings are limited.

The RAID controllers are getting the smarts, and are learning to speak
iSCSI for such purposes as cross-site replication.  Cluster
filesystems on hosts, slowly, too. Perhaps someone figures they can
get even better cost performance on the local backend interconnect
using ata-over-ethernet (with gig-e jumbo frames at least, I hope), by
doing away with many aspects of iSCSI that don't really apply behind
the raid controller.

It might also make sense for low-cost compute clusters and blade
server type environments, I guess.

> I wouldn't go so far as to say it's a good idea.

I tend to agree, but it all depends on what someone makes of it. I can
see opportunities there, and hopefully if it does nothing else it
provides competition to iSCSI to make it realise some of its price
promises.

--
Dan.

--Bzq2cJcN05fcPrs+
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (NetBSD)

iD8DBQFCEU3gEAVxvV4N66cRAkxEAKC1JId6jyyX0GBuh8vh9XiWVMuVPwCeI+ry
AGU5oxX6Fw7qgmK+JEWMif4=
=J36s
-----END PGP SIGNATURE-----

--Bzq2cJcN05fcPrs+--