Enterprise-class ATA Drives

dfung writes "This has been mindlessly discussed many times before here, but Western Digital has now introduced real enterprise-class ATA drives with SCSI-like performance specs and 30% lower price. So now you can buy a real 10K rpm ATA drive. Interestingly enough, they mention the reason for the traditional difference in price between ATA and SCSI which I never have seen mentioned here - it has to do with testing costs, not controller electronics|platter quality|etc. Another interesting tidbit is that 160 million ATA drives were sold last year. I saw about 2 million of them stacked up in the aisles at Fry's Electronics yesterday, but that sure is a lot of drives."

Comments

Hmmm

[IanBevan]

Now I just need some enterprise class pr0n to store one of these suckers...

Re:Hmmm

[kingtonm]

Here ya go, why don't you start with this, some major, enterprise server porn. http://www.sun.com/smrc/photos-sun/downloads/datac enter-gen03s.jpg

heh; common misconseption

[lingqi]

Interestingly enough, they mention the reason for the traditional difference in price between ATA and SCSI which I never have seen mentioned here - it has to do with testing costs, not controller electronics|platter quality|etc.

being IN the semiconductor test industry, it's really interesting how rarely does people really consider the necessity, and challenges, let alone costs, in testing.

few people realize that, for example (I am saying this example purely based on speculation, but a well-formed one) that the athlon MP chip cost difference is in a large part the extra test they run on it. You see - testing cost money, anything that would make test run longer means that more money has been spent on that part "making" it. One of the things the test industry is always talking about is speeding up testing, as a way to reduce testing costs.

aaanyway... next time anybody look at some nifty / advanced gadget, think to yourself "how the heck do they test THAT?" especially with things that have fast interfaces or embedded components...

anyway. erm - to stay on topic: ATA drives could handle 10k platters; I think the point about scsi has always been the more "industrial scalability / reliability / throughput / whatever" that's the selling point. well, and the fact that back in the day you can't buy IDE CDR drives.

Re:heh; common misconseption

[lingqi]

hmm; don't think of it that way. it's simply not possible for people to go and plug in stuff by hand on a cable and test them. usually. that's wayyyyyyy in the end anyway - intergration testing in software people's terms?

I would *suspect* that scsi chipsets have more things to be tested than ATA ones, since as you may notice, they are supposed to work with 15 devices OR by themselves, possibly providing onboard termination or not.

testing often starts at the wafer stage (where each chip is probed and marked, failed ones are crushed, oftenly), and again when chips are packaged - usually speed sorted / repaird (if possible - a lot of memory devices support repair) at this time. After that, integration testing is actually EASIER because this is when you have a whole set of firmware commands to work with, etc.

Frequently chips have dedicated testing commands, though (that you don't get to know), so things are not completely dire. most flash memory have test modes, for example, where if you put in a code sequence it will write the entire array into, say, a checker board pattern. This is to avoid massive delays of half microsecond writing each location, sequentially. Logic chips (like, say, scsi chipsets) usually have a different challenge - they have embedded subsections, often cache, that you don't have access to directly.

now, to get "into the chip" you will have to sequentially put in the test patterns / vectors into special registers that reside on the lines that run between each embedded component. one register at a time (usually sequentially through a few (dozen or less) pins. testing is expensive, but pins more so ;). after each "scan burst" you toggle the clock, and sequentially read out all the registers to see if the chip did what you wanted it to do.

this gets back to the scsi being harder to test - probably the control chipsets are more complex. I can't imagine the mechanical sections being any different (besides the 15krpm ones, anyhow) - generally when something have to communicate with a bunch of other things (like scsi) versus just a few (ATA), the former is more pain in the butt testing wise.

oh, btw - more PIN is also another factor to costs. testers have a limited number of pins, so if you have more pins to test, you test less per turn. can't speak authoratatively on the pincount of drive controller chipsets... just FYI here.

side note: one thing you realize after being in testing is that semiconductor manufacturs often (or, sometimes - depending on the manufacture) puts a LOT of margin into their chips. when they say the chip is rated 75 degrees C, they really mean 75 degrees because the chip was TESTED at that temperature.

ok. long rant... gotta stop now.

Warp 10, scotty!!

[patrixmyth]

Enterprise, class, eh? I just can't resist.

Kirk: Scotty, give me 10,000 rpm on those ATA drives!

Scotty: Captain, she can't take it!

Kirk: Damn it, Scotty, you.... promised me.... SCSI speeds!

Anywho, forget about Enterprise Class ATA Drives, when do I get a tricorder, or at least voice recognition built into my five-button wireless optical mouse?

Re:30%?

[hxnwix]

SATA offers all the speed benefits of SCSI (such as command queueing and device initiated data transfer). In addition, it is one drive per channel. But "wait," you say, "servers need lots of drives in raid5 on each channel!!!" One drive per channel is a blessing in disguise.

From time to time I've seen drive logic fail (as opposed to surface errors), which often brings down the entire SCSI channel. With raid5, you can only afford to lose one drive and perhaps a couple hot spares. Certainly not 14 drives in one shot. SCSI is many pinned, and SCSI raid adapters are designed to have many drives on each channel. One drive per interface is extremely costly and impractical. In this respect, SATA is more robust.

"If one drive per channel serial interfaces are so good, why weren't the used in the first place," you might wonder. Modern high clock rate microcontroller technology permits much higher frequency twisted pair serial interfaces that can offer superior bandwidth to older parallel, ribbon cable interfaces. If SCSI were being designed today it would look something like firewire, which I'm sure you're not biased against. Don't be fooled by the ATA moniker.

Enterprise?

[StealthSock]

It's been a long road
Gettin' from there to here
It's been a long time
But a fast ATA is finally here
I can download pr0n really fast at last
So much that I'll go blind
Slow ATA's not gonna bottleneck no more
No it's not gonna change my mind
'Cause I've got pr0n, lots of pr0n
I've got so much I dont have to
Ever leave the house
Thanks to faster ATA
I've got such hairy palms
Because of my fast hard drive
I've got pictures
Of all the pr0n stars
I've got (I've got) I've got (I've got) I've got
Pr0n
Lot's of pr0n

Re:30%?

[Lumpy]

Bingo!

Let's see you have 14 drives on a single IDE chain and then do a copy between drives.

Or how about the simple fact that you can get SCSI Ultra 360 that are nearly 3 times faster than anything you can buy that is IDE.

Or the fact that My SCSI drives come with 5 year warranty's The only SCSI drive I have ever had fail are reallllllly old. and EVERY scsi drive I have in service (over 120 of them) haven't been spun down or sat idle for over 4 years.

The new IDE might be close, but until they get proof of reliability under their belt like SCSI has It's only a watch and see item.

SCSI is known to be bullet proof and faster. enterprise ATA is not. so the next 5 years they had better not pull an IBM and produce the worlds crappiest drives.

SCSI is great but...

[gklinger]

Why aren't SCSI drives available in the same densities as ATA drives?

The biggest SCSI drives I've seen are just less than 150Gb but Maxtor makes a 250Gb ATA drive. Is there a technical reason why there isn't size parity?

I've had a preference for SCSI drives for years and I've come to accept that I have to pay a steep premium (and now I know why) but what frustrates me is the density, or lack thereof, with SCSI drives.

Why so difficult?

[Anonymous Coward]

It is always interesting to see all the jumping through hoops to defend the use of SCSI. But I think it is all bullshit.

Every manufacturer could, at any time, start producing a diskdrive that has the mechanical and head/servo electronics of an existing SCSI drive integrated with an ATA bus interface. It would have the reliability of the SCSI drive, and assuming that manufacturer has experience in ATA electronics there is no reason to assume that it would have problems on that end.

No need to have it in the market for 5 years to prove reliability. Disk drives are not even in the market for such a long time.

No, they just want to sepatate two different price categories and don't want to blur that gap by offering drives with features from both sides.

This is a Serial ATA drive

[necere]

This is a Serial ATA drive, which the article even mentions (second paragraph: "...Enterprise Serial advanced technology attachment..."), but then proceeds to call it an ATA drive (instead of SATA) for the rest of the article.

Here's a somewhat less misleading article.

Re:Your looooooong reply..

[lingqi]

i believe you are thinking of different things.

chips and tech becomes mature and their FAILURE RATE decreases. mature technology does not cost less to test. On the whole SCSI is still a more complex technology, and I would not be surprised if tested with higher margin / more thoroughly due to the "enterprise level reliability" thing.

besides, as devices gets more complex and more "mature," generally the testing costs increase because you have all these new features, plus the old features, plus the shit that keeps it backwards compatible, to test. you can do better on the profit margin / cost side by making ships that have a lower failure rate, but that does not mean chips gets tested less, or it takes shorter to test them. On the contrary, it usually goes the other way.

Anyhow, example: RAMBUS was expensive because it was a "cutting edge" manufacturing process. the output impedence of the chips had to be controled very precisely, which is difficult to do and a lot of it failed at test - driving up the cost. as process matured, less failed and price came down. but each chip still went through the same routine, and sat the same amount of time on the testers* and took the same number of pin-capacities**, so the TESTING COST stays the same***.

* as memory size increase, they sit longer, usually
** similarly, wider buses takes more pins
*** so in the end testing cost usually increases.

separate the two concepts.

MaximumPC did a IDE vs. SCSI

[Jethro On Deathrow]

This month's issue pits IBM's best IDE vs. a Seagate Cheetah SCSI.

The Winner? The SCSI drive by a margin of more than 30%. There is still a huge difference, especially in the random seek and file transfer areas.

Enterprise-class??

[1s44c]

Enterprise-class??

Great, I'll bear it in mind if I ever build a starship.

Re:Seems like a good idea

[Anonymous Coward]

Or just buy yourself an inexpensive 3Ware IDE RAID controller. The RAID-1 two disk controller is only about $120. The RAID-5 supporting 4 channel one is around $400 if I remember right. Considering they have built in Linux support and they have open source drivers we should really help support this company because companies like this are few and far between. I took my RAID-1 controller out of the box, popped it in my new system, put two "special edition" WD 80GB hard drives on it, created the mirror in the card's setup, and booted Linux and it recognized it as a SCSI controller with a SCSI disk attached. Just seems like a cleaner solution than mucking around with software raid. I never did like software raid much.

Western Digital Support Outsourced, Staff Laid Off

[gottabeme]

In my Ask Slashdot article asking about how heat and vibration affect hard drives, a WD tech support staffer responded, and I corresponded with him via e-mail. He told me that he was being laid off. Here are the details, quoted with permission:

Regarding tech support, basically our division manager decided he could cut costs hugely if he closed our location and contracted the work out outside the country. This may be, but people who don't have a clue about the product they're supporting and can just read the script in front of them is NOT one of the reasons we've won like every tech support award in the industry for the past 7 years. (FYI, our site opened 7 years ago...coincidence?) Look, I don't like to give companies a bad name, even the one that's laying me off. They still make a damn fine product, even if some of the execs have their head so far up their asses you'd have to send in a mining rescue team to find it. But indeed, don't count on good tech support after Feb 13. Our division manager has announced it's his goal to handle 95% of support with people reading off of a basic script with no training. Anything else above a replacement call or the most basic installation you'll have to pay for, and the people who know what we're doing (us) won't be around any longer even. Anyway I should stop about that, I tend to get a bit ranty at times :)