ATA Reliability: Seagate, WD, Maxtor

flux writes:
Aha. Everyone else is describing high volume storage servers, not
desktops.

Which by everyone's description are just as prone to failure. Otherwise,
why would one have redundant components?

They're less prone to failure but they have heavier workloads and the
consequences of failure are higher.

On Sat, 16 Apr 2005 19:16:32 GMT, flux wrote:

In article 1113587080.671422@smirk,
(Ralph Becker-Szendy) wrote:

The smallest computer room of mine in the last 8 or so years had 6
racks worth of equipment. The largest has maybe 40 or 50 racks (I
haven't counted, since it is shared with a few other people, I know
which racks are mine). This doesn't count a large computer room where
I was only in charge of a small part (my part was 4 racks, but the
whole room must have had several hundred racks worth, as big as 2
football fields; rumor has it that it had over 3000 servers in it).

I wonder how else of this is just rumor.

Try connecting a few hundred disks in 4 or 5 racks with SATA cables.

Why would you need ANY SATA cables to connect them? There is this thing
called a backplane. Surely, there must be a few of these in that big
computer room of yours.

Because it's point to point stupid. If there are, say 7 disks
connected to a normal backplane you need to connect 7 cables to the
backplane. That's the purpose of the Serial ATA II Port Multiplier
Technology which for all intensive purposes is vaporware or at least
immature.

On Sat, 16 Apr 2005 21:15:00 GMT, Curious George wrote:

Try connecting a few hundred disks in 4 or 5 racks with SATA cables.

Why would you need ANY SATA cables to connect them? There is this thing
called a backplane. Surely, there must be a few of these in that big
computer room of yours.

Because it's point to point stupid. If there are, say 7 disks
connected to a normal backplane you need to connect 7 cables to the
backplane. That's the purpose of the Serial ATA II Port Multiplier
Technology which for all intensive purposes is vaporware or at least
immature.

& one more thing- you also have to connect the anode & diode wires for
the drive activity from the controller to the backplane. It gets so
messy that some of the larger controllers don't even bother with
individual drive activity.

This is why presently the external soho boxes group the drives into a
single logical unit and connect to the computer on one SATA channel -
but that means max bandwidth is now 150 MB/sec minus overhead for all
the drives in the external box in addition to any other limitations/
inefficiencies of this uncommon, low-end solution.

Companies that are trying to make SATA fit better for larger
enterprise scenario make boxes with a SATA-SCSI bridge. These are
not bargain basement/soho cheap & you're still using scsi cabling &
HBA's so the economics are somewhat questionable. Very questionable
if using Raptors.

In short 50 SATA spindles per computer is indeed a mess & not as cheap
as you think.

Curious George wrote:

Companies that are trying to make SATA fit better for larger
enterprise scenario make boxes with a SATA-SCSI bridge. These are
not bargain basement/soho cheap & you're still using scsi cabling &
HBA's so the economics are somewhat questionable. Very questionable
if using Raptors.

This is just as effective and just as foolish as putting a 3.5HP Brigs &
Stratton lawnmower engine in a Porsche 911. If you want a system to be like
SCSI you should just buy SCSI in the first place. All this dicking around
doesn't change the fact that all the wasted money and time you threw at SATA
will ever get you to square one.

In short 50 SATA spindles per computer is indeed a mess & not as cheap
as you think.

Why screw with them in the first place? I can see if you're building a
novelty toy gaming system for your teenage children you might have
something, but deploying this crap in an enterprise environment is totally
and utterly foolish.



Rita

Curious George wrote:

Because it's point to point stupid. If there are, say 7 disks
connected to a normal backplane you need to connect 7 cables to the
backplane. That's the purpose of the Serial ATA II Port Multiplier
Technology which for all intensive purposes is vaporware or at least
immature.

All of which renders it unpractical and useless in real world applications,
except, of course, in AMD boxes, which nobody real cares about anyway. SCSI
equals one cable per backplane. The options are infinite with SCSI.



Rita

In article ,
Curious George wrote:

On Sat, 16 Apr 2005 19:16:32 GMT, flux wrote:

In article 1113587080.671422@smirk,
(Ralph Becker-Szendy) wrote:

The smallest computer room of mine in the last 8 or so years had 6
racks worth of equipment. The largest has maybe 40 or 50 racks (I
haven't counted, since it is shared with a few other people, I know
which racks are mine). This doesn't count a large computer room where
I was only in charge of a small part (my part was 4 racks, but the
whole room must have had several hundred racks worth, as big as 2
football fields; rumor has it that it had over 3000 servers in it).

I wonder how else of this is just rumor.

Try connecting a few hundred disks in 4 or 5 racks with SATA cables.

Why would you need ANY SATA cables to connect them? There is this thing
called a backplane. Surely, there must be a few of these in that big
computer room of yours.

Because it's point to point stupid. If there are, say 7 disks
connected to a normal backplane you need to connect 7 cables to the
backplane. That's the purpose of the Serial ATA II Port Multiplier

That's only if you are connecting the backplane to card. Otherwise, you don't need any cables.

In article ,
Paul Rubin wrote:

flux writes:
Aha. Everyone else is describing high volume storage servers, not
desktops.

Which by everyone's description are just as prone to failure. Otherwise,
why would one have redundant components?

They're less prone to failure but they have heavier workloads and the

There is some generalization here. Some models of equipment such as IBM
eSeries and Infortrend RAID controllers, aren't aren't any better than
desktops and may even be worse.

In article ,
"Rita Ä Berkowitz" ritaberk2O04 @aol.com wrote:

Curious George wrote:

Because it's point to point stupid. If there are, say 7 disks
connected to a normal backplane you need to connect 7 cables to the
backplane. That's the purpose of the Serial ATA II Port Multiplier
Technology which for all intensive purposes is vaporware or at least
immature.

All of which renders it unpractical and useless in real world applications,
except, of course, in AMD boxes, which nobody real cares about anyway. SCSI

In other words, they are rapidly becoming popular.

In article ,
Curious George wrote:

& one more thing- you also have to connect the anode & diode wires for
the drive activity from the controller to the backplane. It gets so
messy that some of the larger controllers don't even bother with
individual drive activity.

That's true and it's just dumb.

This is why presently the external soho boxes group the drives into a
single logical unit and connect to the computer on one SATA channel -
but that means max bandwidth is now 150 MB/sec minus overhead for all

SOHO boxes? Single channels across the whole box? Who's doing that?

the drives in the external box in addition to any other limitations/
inefficiencies of this uncommon, low-end solution.

Companies that are trying to make SATA fit better for larger
enterprise scenario make boxes with a SATA-SCSI bridge. These are
not bargain basement/soho cheap & you're still using scsi cabling &
HBA's so the economics are somewhat questionable. Very questionable
if using Raptors.

The economics aren't questionable if you want space. For that, you can't beat Hitachi 7K400s.

In short 50 SATA spindles per computer is indeed a mess & not as cheap
as you think.

Yes, it is.

In article ,
"Rita Ä Berkowitz" ritaberk2O04 @aol.com wrote:

Curious George wrote:

Companies that are trying to make SATA fit better for larger
enterprise scenario make boxes with a SATA-SCSI bridge. These are
not bargain basement/soho cheap & you're still using scsi cabling &
HBA's so the economics are somewhat questionable. Very questionable
if using Raptors.

This is just as effective and just as foolish as putting a 3.5HP Brigs &
Stratton lawnmower engine in a Porsche 911. If you want a system to be like
SCSI you should just buy SCSI in the first place. All this dicking around
doesn't change the fact that all the wasted money and time you threw at SATA
will ever get you to square one.

In other words it makes a lot of sense.

In short 50 SATA spindles per computer is indeed a mess & not as cheap
as you think.

Why screw with them in the first place? I can see if you're building a
novelty toy gaming system for your teenage children you might have
something, but deploying this crap in an enterprise environment is totally
and utterly foolish.

So why is everyone doing that?