Beta
×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

MS Researchers Call Moving Server Storage To SSDs a Bad Idea

timothy posted more than 5 years ago | from the professor-if-these-projections-are-accurate dept.

Data Storage 292

An anonymous reader writes "As an IT administrator did you ever think of replacing disks by SSDs? Or using SSDs as an intermediate caching layer? A recent paper by Microsoft researchers provides detailed cost/benefit analysis for several real workloads. The conclusion is that, for a range of typical enterprise workloads, using SSDs makes no sense in the short to medium future. Their price needs to decrease by 3-3000 times for them to make sense. Note that this paper has nothing to do with laptop workloads, for which SSDs probably make more sense (due to SSDs' ruggedness)."

Sorry! There are no comments related to the filter you selected.

Not every tool is right for every application?! (5, Funny)

Tumbleweed (3706) | more than 5 years ago | (#27506519)

News at 11!

Re:Not every tool is right for every application?! (5, Funny)

VeNoM0619 (1058216) | more than 5 years ago | (#27506587)

That's hardly the issue... notice how they say 3-3000 times cheaper. Meaning a $3000 SSD would have to cost $1 for them to consider it... Don't you love pulling numbers of your ass?

Re:Not every tool is right for every application?! (4, Insightful)

AvitarX (172628) | more than 5 years ago | (#27506805)

I actually don't think times cheaper makes any sense.

I hear it all the time, but it is meaningless.

3000 times cheaper than what? The current price?

If I am selling something that is now "twice as cheap" is that half the price?, double the discount?, twice as shoddily made?

Re:Not every tool is right for every application?! (1)

sexconker (1179573) | more than 5 years ago | (#27506915)

You have to define cheap.
A common sense definition puts it as the reciprocal of expensive.

Just like saying twice as slow or twice as fast.
You logically define slow as the time it takes, and fast as the reciprocal.

Re:Not every tool is right for every application?! (2, Insightful)

AvitarX (172628) | more than 5 years ago | (#27507205)

Funny, my other complaint is twice as slow.

The problem I see is 3 times slower doesn't multiply anything by 3, it divides it.

and intuitively slower of cheaper are not inverse, since we use statements like 5 dollars cheaper.

Re:Not every tool is right for every application?! (1)

geminidomino (614729) | more than 5 years ago | (#27507285)

The problem I see is 3 times slower doesn't multiply anything by 3, it divides it.

It divides velocity by 3, but it also multiplies time by 3.

Re:Not every tool is right for every application?! (-1, Flamebait)

Razalhague (1497249) | more than 5 years ago | (#27506929)

"x times cheaper" = current-price - (x * current-price). Duh.

Re:Not every tool is right for every application?! (3, Funny)

jebrew (1101907) | more than 5 years ago | (#27507079)

So if it's $100, and they say 3 times cheaper, then it's got to be:

$100 - (3 * $100) = -$200

Hell, if they pay me $200 AND give me an SSD, I'll be a happy person.

Re:Not every tool is right for every application?! (1)

AKAImBatman (238306) | more than 5 years ago | (#27507111)

Oh, that makes sense. So if an SSD is $20/GB right now, I'd need it to go down by 20 - (3000 * 20).

Or in other words, I'll purchase the SSD when the server manufacturer pays me $59,980/MB.

Hmm... something doesn't seem quite right here...

Re:Not every tool is right for every application?! (-1, Flamebait)

Razalhague (1497249) | more than 5 years ago | (#27507435)

Hmm... something doesn't seem quite right here...

That'd be your sense of humour.

Re:Not every tool is right for every application?! (0)

Anonymous Coward | more than 5 years ago | (#27507521)

Uhhh.. reciprocal means divide instead of multiply - not whatever the hell you're doing. So, 20 * 3000 is 3000 times more expensive, 20 / 3000 is 3000 times cheaper.

So, purchase the SSD when it hits around $0.67/MB

Re:Not every tool is right for every application?! (1)

Lord Pillage (815466) | more than 5 years ago | (#27506955)

Think of the phrase "as cheap" as the inverse of "as expensive". Thus, 2 times cheaper == 1/2 times as expensive (i.e. half price). At least that's how a geek could make sense of it, since you are correct, it doesn't make sense when taken literally.

Re:Not every tool is right for every application?! (0)

Anonymous Coward | more than 5 years ago | (#27507047)

A geek would prefer the much less ambiguous term, "half as expensive." This "twice as cheap" bullshit really gets under my skin even if there is a commonly accepted meaning.

Re:Not every tool is right for every application?! (1)

noidentity (188756) | more than 5 years ago | (#27507549)

If A is X times cheaper than B, we must first calculate the cheapness of B. If B should be priced at P, but is really priced at Q, where Q<P, we can calculate B's absolute cheapness as P-Q, and relative cheapness as (P-Q)/P. Therefore if A should cost P', then an A that is X times cheaper than B will cost P' discounted by the relative cheapness of B multiplied by X, that is, P'-P'X(P-Q)/P. Quite simple really.

Re:Not every tool is right for every application?! (5, Funny)

Cormacus (976625) | more than 5 years ago | (#27506753)

I dunno about that. I'm pretty sure that if your only tool is a hammer, all of your problems start looking like nails . . . allowing the hammer to be "applied" to every application . . .

Re:Not every tool is right for every application?! (2, Funny)

Tumbleweed (3706) | more than 5 years ago | (#27507015)

I dunno about that. I'm pretty sure that if your only tool is a hammer, all of your problems start looking like nails . . . allowing the hammer to be "applied" to every application . . .

I think an SSD would make for a very expensive hammer. Still, think about the low latency of such a hammer! Plus with the wear levelling feature, the useful life of and SSD hammer seems like it would be much more reliable over a spinning disc hammer. And the lower power requirements could pay for itself very quickly if you have an entire server room of carpenters. I don't think they did the math right on this one.

How about a policy: NO PAYWALLS! (5, Insightful)

nweaver (113078) | more than 5 years ago | (#27506543)

This is an ACM article behind a paywall.

How about a slashdot policy of not linking to articles behind paywalls?

Re:How about a policy: NO PAYWALLS! (0)

Anonymous Coward | more than 5 years ago | (#27506565)

Link away, a little bit of accidental vamping can't help sites that use the flawed micropayment model.

Free Link: Google HTML Cache (2, Informative)

A. B3ttik (1344591) | more than 5 years ago | (#27506879)

That wasn't hard. [74.125.47.132]

Re:How about a policy: NO PAYWALLS! (-1)

Anonymous Coward | more than 5 years ago | (#27506569)

This would require the editors to actually click the link.

Re:How about a policy: NO PAYWALLS! (0)

Anonymous Coward | more than 5 years ago | (#27506575)

I have access to the article and this is relevant to my interests, you insensitive clod!!

Wait, no, I don't, carry on.

Re:How about a policy: NO PAYWALLS! (5, Informative)

Timothy Brownawell (627747) | more than 5 years ago | (#27506603)

How about a slashdot policy of not linking to articles behind paywalls?

Seriously, it's even worse than the "free registration required" links that we used to have problems with.

Original PDF at http://research.microsoft.com/pubs/76522/tr-2008-169.pdf [microsoft.com] .

Re:How about a policy: NO PAYWALLS! (1)

jacksinn (1136829) | more than 5 years ago | (#27506609)

I'm an ACM member and don't have access. Looks like I would have to shell out another $99 per year in addition to my membership fee for the digital library.

Re:How about a policy: NO PAYWALLS! (1)

Knackered (311164) | more than 5 years ago | (#27506921)

I'm an ACM member an I do have Digital Library access. It's worth $200 a year to me to get access to years of research, including cutting-edge stuff, that may help me do my job better. It's also tax deductible as a professional subscription.

Re:How about a policy: NO PAYWALLS! (1)

Enokcc (1500439) | more than 5 years ago | (#27506667)

How about a slashdot policy of not linking to articles behind paywalls?

And forget about most of the scientific articles?

Re:How about a policy: NO PAYWALLS! (0)

Anonymous Coward | more than 5 years ago | (#27507605)

I'd rather see scientists ditching these dinosaurs altogether. Little encouragement probably wouldn't hurt.

Re:How about a policy: NO PAYWALLS! (0, Offtopic)

phantomcircuit (938963) | more than 5 years ago | (#27506789)

What you mean that UC Berkeley doesn't pay for you to have access to this article?

sucks for you

Re:How about a policy: NO PAYWALLS! (1)

houstonbofh (602064) | more than 5 years ago | (#27507209)

This is an ACM article behind a paywall.

How about a slashdot policy of not linking to articles behind paywalls?

Would love to know if someone notices the logs. 10 bazillion hits on the summery page. 5 paid reads. No new accounts... :)

Paid ACM subscription (5, Insightful)

eples (239989) | more than 5 years ago | (#27506555)

Their price needs to decrease by 3-3000 times for them to make sense.

Hm. I was thinking the same thing about the ACM subscription.

Re:Paid ACM subscription (1)

kschendel (644489) | more than 5 years ago | (#27506833)

Indeed. That was the conclusion I came to a few years back.

It might be different if one could eliminate CACM as part of the deal, as CACM became completely worthless about 20 years ago. I asked, and was told that I could suppress the CACM subscription ... for zero reduction in dues. Whatever.

Re:Paid ACM subscription (1)

geminidomino (614729) | more than 5 years ago | (#27507399)

Ditto. I got a year free for being inducted into UPE (Smart move on their part, since I've got fsckall time to read anything between work and studies), but a quick scan of it told me that it wasn't worth $200 (or even the $99 to renew just membership)

 

they already cost less per gig than some SAS drive (5, Insightful)

hxnwix (652290) | more than 5 years ago | (#27506567)

SSD is already cheaper per gig than some SAS drives. Also, 3-3000 times? What the hell sort of estimate is that?

Re:they already cost less per gig than some SAS dr (5, Insightful)

Larry Clotter (1527741) | more than 5 years ago | (#27506627)

It's called "pulling numbers out of your ass".

Re:they already cost less per gig than some SAS dr (5, Funny)

Yvan256 (722131) | more than 5 years ago | (#27506739)

What do you mean, an african or european ass?

Re:they already cost less per gig than some SAS dr (0)

Anonymous Coward | more than 5 years ago | (#27506867)

I lolled. Thanks.

Re:they already cost less per gig than some SAS dr (1)

Daimanta (1140543) | more than 5 years ago | (#27507087)

I don't know!

waaarrggghhhh

Re:they already cost less per gig than some SAS dr (1)

nih (411096) | more than 5 years ago | (#27506913)

I haven't got an ass you insensitive clod!

Re:they already cost less per gig than some SAS dr (1)

jhol13 (1087781) | more than 5 years ago | (#27506993)

I'd say Microsoft does not have any "solution" in which SSD would help.

OTOH Sun does ... http://www.sun.com/storage/flash/ [sun.com] . YMMV.

Re:they already cost less per gig than some SAS dr (0)

Anonymous Coward | more than 5 years ago | (#27507249)

I'd say Microsoft does not have any "solution" in which SSD would help.

ReadyBoost in Vista is doing basically the same thing (at least for reads). So the foundation is already there (and has been for a couple of years now).

Re:they already cost less per gig than some SAS dr (1)

Penguinoflight (517245) | more than 5 years ago | (#27506775)

I can only guess they're referring to differently priced SSDs. Some cost in the thousands, but provide top-teir performance. Their price would be justified at approximately 1/3rd the current price, as that's what would be necessary to provide similar cost/performance to a raid array of rotational drives.

On the other hand, the low cost MLC ssds typically provide lower performance than a single rotational drive at a cost premium in the range of about 100x the cost of a rotational drive. These lower cost drives are frequently seen as needing an improvement to capacity to be worth considering, you see them in 64gb sizes and the like. They would have to be available at a price lower than rotational drives in similar capacities to be considered, which would a 100+ scale price reduction.

I'm not seeing the 3000x reduction necessary, I guess a small exaggeration for effect may be the source of that.

Re:they already cost less per gig than some SAS dr (1)

drsmithy (35869) | more than 5 years ago | (#27507293)

I can only guess they're referring to differently priced SSDs. Some cost in the thousands, but provide top-teir performance. Their price would be justified at approximately 1/3rd the current price, as that's what would be necessary to provide similar cost/performance to a raid array of rotational drives.

The interesting thing is, according to the performance table on page 6, the SSD they used only had write performance of ~350 IOPS. Either that number is missing a zero, or something is _seriously_ wrong with their SSD.

If it's the latter, then clearly any conclusions drawn from that write performance are completely invalid.

Re:they already cost less per gig than some SAS dr (1)

orclevegam (940336) | more than 5 years ago | (#27506807)

SSD gives phenomenal random read performance, equally good serial read performance, and average write and random write performance (at least if you get a good SSD, the low end ones using the crap IO chips are worse than budget HDs). The only way to beat the read performance of a good SSD is a really expensive SAS RAID, and even then it's not going to be by much. Yes you can take a hit on serial write performance, but not much of one (it's on par with most medium to high end HDs, with surprisingly few high end HDs able to outperform it). If you're primarily going to be doing reads, particular random reads, or even if you're going to be doing mostly random writes rather than serial writes, an SSD is probably a good idea.

Re:they already cost less per gig than some SAS dr (1)

sexconker (1179573) | more than 5 years ago | (#27507113)

High-end HDDs still edge out SSDs for serial reads in many setups.

Keep in mind that write performance degrades over time (goes from great to very good) as the pages get full.

When you're out of free pages, you have to read an entire block of pages to cache, erase the entire block, then write back the new block.

Current OSs and controllers do not yet support the "yes, actually delete it" command, and current controllers do not yet support any sort of automatic drive-level page consolidation.

If money is no object, then SSD is a good choice. A better choice is a RAM drive, though prices for these make SSDs look like a bargain.

$-for-$ there is no contest. HDDs win. I think they've got about 2 years left before they start to be marginalized to bulk storage devices.

Re:they already cost less per gig than some SAS dr (1)

vrmlguy (120854) | more than 5 years ago | (#27507301)

If you're primarily going to be doing reads, particular random reads, or even if you're going to be doing mostly random writes rather than serial writes, an SSD is probably a good idea.

Which, as you undoubtedly know, is why enterprise data centers use SANs (Storage Attached Networks) with multiple tiers of storage. You use big slow drives for archival storage (old emails, for instance), and smaller faster drives for day-to-day use (databases, etc). Flash drives get used when performance really matters, such as database indexes, not the actual data.

Re:they already cost less per gig than some SAS dr (2, Interesting)

Znork (31774) | more than 5 years ago | (#27506819)

Of course, SAS drives are also often too expensive to survive a purely cost/benefit driven analysis. For many real-world loads you're better off adding more spindles which can give you similar iops per dollar but with the added benefit of vastly more storage space.

There's a lot of snake oil and very little quality analysis in enterprise storage these days, so it's good to see at least some do attempt to do actual real-world cost/benefit calculations before jumping onto the marketing train.

Re:they already cost less per gig than some SAS dr (1)

darury (1235658) | more than 5 years ago | (#27506969)

For many real-world loads you're better off adding more spindles which can give you similar iops per dollar but with the added benefit of vastly more storage space.

I'm not going to dispute the half of adding more spindles for better performance, but additional storage space is not always an added benefit. Do you have any idea how frustrating it can be to explain to users that "yes, we have 4TB of unused space" and "no, you can't put another database there without killing your existing one"?

Re:they already cost less per gig than some SAS dr (1)

vrmlguy (120854) | more than 5 years ago | (#27507425)

I'm not going to dispute the half of adding more spindles for better performance, but additional storage space is not always an added benefit. Do you have any idea how frustrating it can be to explain to users that "yes, we have 4TB of unused space" and "no, you can't put another database there without killing your existing one"?

That's called short stroking. No, really, you only use the innermost tracks of your drives to avoid long seeks, turning a 144 GB drive into 72 or even 36 GB. If you do it right, the unused capacity doesn't show up on any of your reports, so no one ever realizes that the extra space is there.

Re:they already cost less per gig than some SAS dr (1)

drsmithy (35869) | more than 5 years ago | (#27507055)

Of course, SAS drives are also often too expensive to survive a purely cost/benefit driven analysis. For many real-world loads you're better off adding more spindles which can give you similar iops per dollar but with the added benefit of vastly more storage space.

You need 2-3x as many SATA drives as 15k SAS/FC drives to get equivalent IOPS. That means 2-3x as much physical space required, probably around 1.5-2x as power usage and decreased reliability overall.

Storage volume is rarely a concern when the primary objective is performance.

Re:they already cost less per gig than some SAS dr (1)

Carewolf (581105) | more than 5 years ago | (#27507253)

probably around 1.5-2x as power usage and decreased reliability overall.

Have you ever heard of thing called RAID? It is actually designed for inexpensive crappy of the shelf disks.

Re:they already cost less per gig than some SAS dr (1)

drsmithy (35869) | more than 5 years ago | (#27507545)

Have you ever heard of thing called RAID? It is actually designed for inexpensive crappy of the shelf disks.

More disks means less reliability due to more points of failure. A 16-drive array is more likely to have a failure than a 6-drive array by virtue of simple statistics.

Re:they already cost less per gig than some SAS dr (1)

Z00L00K (682162) | more than 5 years ago | (#27506861)

It depends on the application which to select right now, but in the long run the SSD drives will have an advantage.

So soon we may no longer need those noisy hard disks at all.

And when a storage is built on flash memories it may be possible to work with it in segments where parts of the disk isn't powered in order to save power and generate less heat. The latter is a huge advantage in datacenters where cooling is expensive.

The ruggedness is also an advantage, but not in datacenters. What you usually want in a datacenter is good performance to a low cost, but you also want reliability.

Re:they already cost less per gig than some SAS dr (1)

afidel (530433) | more than 5 years ago | (#27506959)

Not ones you'd use in an enterprise! X-25e is the only SLC based flash with a decent controller under $1k and it's still $24/GB. Unless you have a WORM application that needs fast seeks (pretty rare) MLC based flash isn't a good fit for most enterprise applications. The only areas we've found for them are log drives for high transation database servers where the insane IOPS per $ make sense and cache for a BI system which still sees enough writes to rule out MLC. Oh and their analysis is based on a rather small set of data, I'm at a midsized shop and I have more storage online and a more varied workload than that.

Re:they already cost less per gig than some SAS dr (1)

C. E. Sum (1065) | more than 5 years ago | (#27507157)

More like 13$/GB. HTH! HAND!

http://www.provantage.com/intel-ssdsa2sh064g101~7ITE90J5.htm [provantage.com]

X25E SLC 64GB 2.5INCH SATA SSD $827

And Provantage is rarely a price-leader.

Re:they already cost less per gig than some SAS dr (1)

afidel (530433) | more than 5 years ago | (#27507405)

Cool, when I bought mine the 64GB wasn't available and the 32GB was almost that much. That's why SSD's are so cool right now, even at the high end the $/GB is falling rapidly, much more so then enterprise HDD's. 450GB 15K FC drives are about three times that much so about 1/3rds the $/GB but MUCH higher $/IOP.

Re:they already cost less per gig than some SAS dr (1)

qoncept (599709) | more than 5 years ago | (#27507031)

What is this, an SSD for ants???

Re:they already cost less per gig than some SAS dr (1)

houstonbofh (602064) | more than 5 years ago | (#27507267)

SSD is already cheaper per gig than some SAS drives. Also, 3-3000 times? What the hell sort of estimate is that?

I remember an article a while ago talking about how Windows disk drivers are not optimized for SSD. Now there is a white paper showing how SSD is not practical by Microsoft. So to answer your question, it is a PR estimate.

What if... (3, Insightful)

Thelasko (1196535) | more than 5 years ago | (#27506577)

they don't use NTFS?

HERETIC! (0)

Anonymous Coward | more than 5 years ago | (#27506617)

they don't use NTFS?

I cast thee OUT!

Re:What if... (5, Funny)

Anarke_Incarnate (733529) | more than 5 years ago | (#27506673)

FAT chance.........

Re:What if... (5, Funny)

Rayeth (1335201) | more than 5 years ago | (#27507327)

ext-remely unlikely.

3 to 3000 percent? (3, Insightful)

erroneus (253617) | more than 5 years ago | (#27506643)

My goodness! They have really done their research in order to produce data as accurate as that!

The fact was, they said the same thing when it came to magnetic tape versus magnetic disks. These days, hard drives are cheaper than tapes and will hold their data longer and more compatibly.

Microsoft fears change that they do not control. If they don't control the changes, someone might write them out of the story.

Re:3 to 3000 percent? (0)

Anonymous Coward | more than 5 years ago | (#27506767)

yeah, because microsoft controls so much of the hardware market. i think you're talking about apple.

stop being such a troll.

Re:3 to 3000 percent? (4, Informative)

lgw (121541) | more than 5 years ago | (#27506939)

These days, hard drives are cheaper than tapes and will hold their data longer and more compatibly.

That's entirely false.

Hard drives are vastly cheaper than tape drives, but enterprise quality tape is stil cheaper than enterprise quality HDDs.

Enterprise tape has a proven 20-year shelf life, no HDD does.

I wrote new commercial software that could (and did) work with IBM's 9-track tape format in 1994, 30 years after it released, and there is still hardware and software in use today that can read that hardware format - 45 years of compatibility. The abstract format - ANSI tape labels - is still in niche use for newly saved data today. DLT format is 25 years old, and while I'm not sure you can buy a new drive that reads the original DLT format, used drives are still easy to come by and you can connect them to new SCSI cards.

How easy is it to read an MFM drive (assuming there are more than 0 in the world that still work)? That format is 30 years old, and it would be a real challenge to find a slot on a modern PC that would take an MFM controller, vastly harder than reading a DLT tape. FAT is also about 30 years old, but disk formats older than that are basically extinct.

Re:3 to 3000 percent? (1)

blueg3 (192743) | more than 5 years ago | (#27506963)

Probably shouldn't take mathematical advice from someone who confuses "3 to 3000 times" with "3 to 3000 percent".

Considering that SSD prices vary and performance and workload situations vary more, it's not surprising that there is a range. It's not even surprising that it's a large range. (For example, "if your workload is closer to the optimal profile, the price needs to decrease by a factor of 3; if the workload is closer to the worst-case profile, the price needs to decrease by a factor of 3000". The only odd thing is the factor of 3 in a three-order-of-magnitude range.

Re: 3 to 3000 percent? nope. RTFA. +tape vs disk (1)

danpritts (54685) | more than 5 years ago | (#27507097)

First, they don't say 3-3000 percent (nor do they say 3-3000 times, which is what the original post above says). They say 1-3 orders of magnitude.

re: tape vs. disk:

cheaper per byte, sure.

more compatibly, possibly. i'm not sure whether you'd have an easier time reading a 1985 disk or a 1985 tape; in either case you'd need to do some digging to find the appropriate hardware to read the media.

longer? Nope. That 1985 disk might be readable by ontrack in the lab, but odds are pretty good that it won't work right once you dig up that MFM controller. The tape will probably be fine.

re: TFA:

If you RTFA they do some reasonable analysis.

They ignore the possibility that you might drop the RAID1 on your boot disk and go with a single SSD; I'm certainly considering that and if you think hardware raid controller + a pair of disks vs. 1 SSD the cost is very favorable.

They don't appear to take read latency into account; IOPS and latency are not the same thing.

Also, prices have already fallen significantly since their published data. They list a 32G SSD at $739; the same one is $449 today at http://rocketdisk.com/index.php?cPath=8&gclid=CJbK8OH14ZkCFSQeDQodaikyXA [rocketdisk.com]

Re:3 to 3000 percent? (1)

kenp2002 (545495) | more than 5 years ago | (#27507365)

fud mod parent troll please, not kid born in the 90s should never talk about the tape wars unless they were there...

How much is that (-1, Flamebait)

Anonymous Coward | more than 5 years ago | (#27506685)

Now, I'm no "big city" mathematician, but...

3-3000=-2997

The prices need to drop by negative two thousand nine hundred ninety seven? I guess they did their study using Microsoft Excel :)

GOOGLE HTML CACHE (0)

Anonymous Coward | more than 5 years ago | (#27506705)

Geez, that was hard. [74.125.47.132] Also has a PDF version if you don't like the colors.

XServe (1, Flamebait)

Yvan256 (722131) | more than 5 years ago | (#27506721)

I bet this has absolutely nothing to do with the fact that Apple updated their Xserve yesterday with an SSD option.

From what I understand, the SSD is for the OS itself and not the data/storage.

Re:XServe (4, Informative)

DAldredge (2353) | more than 5 years ago | (#27506905)

Page 1 Microsoft Research Ltd. Technical Report MSR-TR-2008-169, November 2008 Not a thing to do with it.

Re:XServe (2, Informative)

Yvan256 (722131) | more than 5 years ago | (#27507001)

Since when are we supposed to read the articles?

Re:XServe (1)

fuzzyfuzzyfungus (1223518) | more than 5 years ago | (#27507033)

I suspect that it, indeed, doesn't. First of all, writing, editing, formatting, and publishing a paper of any length, even a crap one, usually takes more than 24 hours(college students excepted). Second, and most importantly: many, if not most, server vendors already offer some sort of SSD related options on their hardware. Dedicated Big Serious Storage vendors offer even more, for connection to those servers. Virtually all of those offerings support Windows, and have for some time.

MS has reason to fear, and attempt to respond to, Apple on the desktop; but far less so on the server. Even if they did, Apple is actually among the later server vendors to offer an SSD option, and their option is among the more limited. Nice of them to include it; but barely relevant.

Now, if MS research produced a paper "Filesystems that aren't NTFS cause SSDs to suck, wear out horribly" then that might constitute response to their enemies, rather than a more or less dry analysis of tech trends.

Re:XServe (1)

kenp2002 (545495) | more than 5 years ago | (#27507439)

FUD, the paper was written last year....

Nut-jobs (0)

Anonymous Coward | more than 5 years ago | (#27506755)

This report brought to you by the same people that said Linux was a more expensive platform than Windows. Seriously, why do we even listen to these nutbags?

3-3000 times? (3, Funny)

dAzED1 (33635) | more than 5 years ago | (#27506769)

seriously? "we don't have enough people here. we need between 2-2000 times as many people in the configuration department." Does that sound like I have ANY idea how many people we need?

Sorry, that is a *ridiculous* range to give.

Re:3-3000 times? (1)

Slightly Askew (638918) | more than 5 years ago | (#27506937)

For consumer grade telescopes to be able to see the flag planted on the moon, they would need to be 2-2000 times more powerful.

Makes perfect sense if the item you are talking about varies greatly from one vendors product to the next.

Re:3-3000 times? (1)

dAzED1 (33635) | more than 5 years ago | (#27506987)

1) any telescope that needs to be 2000x more powerful to see a flag planted on the moon isn't a telescope.
2) there is not 1000x the price variance (since what was being discussed was price) between viable SDD vendors for similar products.

Re:3-3000 times? (1)

DaveV1.0 (203135) | more than 5 years ago | (#27507171)

there is not 1000x the price variance (since what was being discussed was price) between viable SDD vendors for similar products.

There may, however, be a 1000x total lifetime cost differential between an SSD solution and a standard HD solution, especially in write-intensive, redundant storage applications.

Re:3-3000 times? (1)

dAzED1 (33635) | more than 5 years ago | (#27507441)

I think you're missing the point. If there's 1000x the total lifetime cost differential, then great! You've got an idea of the difference there.

That has nothing to do with my post, though. To say "2000-3000x difference" would at least be reasonable. "5-10 times difference" might be as well. But when the range of differential goes from a single digit (3) to a number 1000x larger (3000) then the range of potential differential demonstrates - when the specifics of something can be so easily known, like in this case - that your study was crap. The range of differential is ridiculous.

There are only so many SDD configurations that make sense. To do this study, you have to first determine what those configurations might be - that would be part of your postulate. Others can then look at the work of these "researchers" and discredit the findings based on the postulates themselves. With something as discrete, and easily quantifiable as IO usage (no ethical questions of animal testing, no grey areas where we don't really understand what's going on, etc), there is absolutely, positively, no reason to give a differential range where the bottom and top of the range are 1000x apart from each other. That is positively unacceptable in an easily quantifiable field like CS. You get exact freaking numbers. there's nothing exact about the range "3-3000x"

Re:3-3000 times? (1)

DaveV1.0 (203135) | more than 5 years ago | (#27507103)

Sorry, that is a *ridiculous* range to give.

Is it now? Different uses give different read/write profiles for the same server configuration. Different server configurations add to the mix. A write-intensive application on a RAID5 system will have a much different cost/benefit analysis than a read-intensive file server using a single drive. Especially when one factors in the fact that the more often one writes to an flash memory based device, the faster that device wears out.

Your little quote should read: "we don't have enough people here. we need between 3-3000 times as many people per department, depending on the department."

Re:3-3000 times? (1)

dAzED1 (33635) | more than 5 years ago | (#27507227)

Their price needs to decrease by 3-3000 times for them to make sense.

There are plenty of configuration options that are possible...that would not make sense. I could make a mirror set of a raid 1+0 set where each component is a raid5 set composed of.... ...yeah, wouldn't make any sense, would it. Which is why I said it is a ridiculous range. We have the specifics of the situation, there's no reason to go trying to large divergent sets; my very point is that the range of divergence is not reasonable given the specifics. So are you saying the range of divergence is reasonable, given the specifics?

Re:3-3000 times? (1)

dAzED1 (33635) | more than 5 years ago | (#27507269)

"there's no reason to go trying to large divergent sets"

should be: "there's no reason to go trying to compare to large divergent sets"

Re:3-3000 times? (1)

DaveV1.0 (203135) | more than 5 years ago | (#27507471)

Did you bother to read the article?

There is a recent post [slashdot.org] that you might want to read.

Re:3-3000 times? (1)

dAzED1 (33635) | more than 5 years ago | (#27507541)

yes. Doesn't change the position. If the answer is "3-10 times cheaper for most configurations, but 2500-3000 times cheaper for small files" then that would be acceptable. ranges can have elements within them. merely saying, however, "3-3000 times" means that there are situations near 100x, some near 245x, some near 1221x, some near 2331x...and all of them would "make sense" per the "Their price needs to decrease by 3-3000 times for them to make sense" quote.

'Real Workloads' (2, Informative)

dchaffey (1354871) | more than 5 years ago | (#27506925)

What a misleading term - I know of companies using Enterprise SSD in production precisely because it's financially sound for them to utilise the ridiculous speed improvement it provides.

Sure, it's not a lot of companies that are using this yet, but as longevity increases with better garbage collection and write-spreading algorithms as well as stabilty and feature set through maturing software and firmware it's closer than you think.

For clarity, the product wasn't SSD behind SATAII, it was FusionIO's PCI devices.

Re:'Real Workloads' (1, Informative)

Anonymous Coward | more than 5 years ago | (#27507203)

In the enterprise space you will also see people using SSDs in large SAN attached storage arrays. In some cases performance requirements can force you to short stroke (or use only a fraction of the capacity of) disks to meet IO per second requirements. Sometimes weighing the cost of hundreds of mostly empty spinning disks vs a few enterprise flash drives can swing the cost in favor of the flash drives. Floorspace and cooling also need to be taken into account in this case.

Inaccurate summary (4, Informative)

chazzf (188092) | more than 5 years ago | (#27506933)

Hat tip to the anon for the Google cache link (http://tinyurl.com/d2py5r). The summary doesn't quote exactly from the paper, which actually said this:

"Our optimization framework is flexible and can be used to design a range of storage hierarchies. When applied to current workloads and prices we find the following in a nutshell: for many enterprise workloads capacity dominates provisioning costs and the current per-gigabyte price of SSDs is between a factor of 3 and 3000 times higher than needed to be cost-effective for full replacement. We find that SSDs can provide some benefit as an intermediate tier for caching and write-ahead logging in a hybrid disk-SSD configuration. Surprisingly, the power savings achieved by SSDs are comparable to power savings from using low-power SATA disks."

Tell that to someone running an OpenStorage SAN... (-1, Troll)

Anonymous Coward | more than 5 years ago | (#27506943)

each SSD you add to your Sun Open Storage you get like 10,000 IOP/sec. for disk reads. If you are running into high io latency on reads you add more of these cache drives. If the drive goes to hell your cache manager does not store any reads on it and goes directly to mechanical disk. Hybrid storage, NetApp is thinking of doing the same thing with the storage products they sell.

Windows 2020 will have the same features as Open Solaris 10, just wait and see. They will be able to use a SSD as a cache reader I swear! This will allow the boot drive to be on a very reliable disk and the read disk be something that is just fast as hell.

Re:Tell that to someone running an OpenStorage SAN (4, Informative)

drsmithy (35869) | more than 5 years ago | (#27507109)

Windows 2020 will have the same features as Open Solaris 10, just wait and see. They will be able to use a SSD as a cache reader I swear!

They could call it... ReadyBoost.

What it really means (5, Funny)

HangingChad (677530) | more than 5 years ago | (#27507017)

Microsoft researchers provides detailed cost/benefit analysis for several real workloads.

If Microsoft researchers report that SSD's are not cost effective storage, it means that Microsoft is not getting any revenue from SSD storage. Or that they're behind on incorporating SSD's into the server stack. Or they caught blind-sided by the trend like they did with netbooks and are now scrambling to explain why they didn't see it coming. Oh, we found that wasn't cost effective, so we didn't incorporate it.

I really miss the days Microsoft had it together. There was a time they were great to work with. Now they seem like the Three Stooges Do IT. SSD, eh? Oh, a wise guy! SMACK! Wo-wo-wo-wo!

This isn't new or startling (0)

Anonymous Coward | more than 5 years ago | (#27507121)

MS isn't making any observations or claims that aren't already well understood by the corporate IT world wide. You don't adopt new technology into a production environment until 1) its failure modes are infrequent, well understood, and able to be mitigated and 2) it is cost effective and/or necessary from a functional or business standpoint to do so. Last time I checked, SSDs were significantly more expensive and significantly lower capacity than conventional HDDs.

Something's wrong (2, Insightful)

drsmithy (35869) | more than 5 years ago | (#27507139)

They list the write IOPS of their "Enterprise SSD" drive as only ~350. That number seems like it's an order of magnitude too low, which would obviously skew the conclusions.

This is probably a reaction to Sun's L2ARC (4, Informative)

kroyd (29866) | more than 5 years ago | (#27507149)

Sun has been making quite a bit of noise in the storage architecture world with their use of SSDs as intermediate cache to improve reading and writing speeds.

http://blogs.sun.com/brendan/entry/test [sun.com] has some background information, and http://blogs.sun.com/brendan/entry/l2arc_screenshots [sun.com] and http://blogs.sun.com/brendan/entry/my_sun_storage_7410_perf [sun.com] has some performance numbers.

Basically, what Sun is claiming is that by adding a SSD cache layer you can improve IOPS by about 5x, for what amounts to a really small amount of money for say a 100tb system. This is being marketed quite heavily by Sun as well. (The numbers look convincing, and the prices for the Sun Storage servers are certainly very competitive, well, compared to say NetApp.)

IMHO this is just a repeat of the well known Microsoft tactic of spreading massive amounts of FUD about any competing technology that you can't reproduce yourself - you'll have to wait until Windows Server 2013 for this.

Boot partition most critical. (2, Informative)

Vellmont (569020) | more than 5 years ago | (#27507211)

Dismissing using SSD because it's only cost effective for the boot partition is a mistake. Anyone who's put together servers before knows the boot partition is critical to the system, and the hardest part to backup. Once you get a system booted, there's a million things you can do to fix it or restore the relevant data. Getting it bootable if the boot partition is toast is much harder.

Re:Boot partition most critical. (1)

geminidomino (614729) | more than 5 years ago | (#27507519)

Shit, these days you can get a boot partition on a 32MB CF/SD card for a server, might have to go all the way up to 128MB for a desktop box... I wonder if they have SD-to-ATAPI adapters...

Provided, of course, that you're not using an OS that thinks the entire fucking UI needs to live there...

Cherry-picked analysis (5, Informative)

David Jao (2759) | more than 5 years ago | (#27507231)

This paper is biased and premature even by the prevailing low standards of typical CS papers. For example, they model SSD failure, but completely ignore mechanical drive failure, which is far more devastating and commonplace. I kid you not:

Since this paper is focused on solid-state storage, and wear is a novel, SSD-specific phenomenon, we include it in our device models. Currently we do not model other failures, such as mechanical failures in disks.

The correct approach to incomplete data is, of course, to gather complete data, and they have no excuse here, because there is PLENTY of data on mechanical drive failure rates. However, if you are not willing to do that, the least you can do is ignore the data equally on both sides. The authors' failure to treat both sides equally leads to a hopelessly biased and skewed analysis.

Read the Paper (5, Informative)

kenp2002 (545495) | more than 5 years ago | (#27507271)

I just finished the reading the paper.

The paper boils down to this:

SSD disk when measured against IOPS, Watts, and Capacity in relation to cost based on several different server types is not cost effective yet. Depending on the type of server costs need to come down at least 3 fold, and under some scenarios as much as 3000 times. Hosting MP3s that are largely sequental, low write storage SSDs are 3000 times over priced. For insaine random IO scenarios that need to come down 3 fold to make it worth it compared to conventional drives.

Depending on the type of server they can perform worse then standard mechanical disks.

They found no advantage to 15k RPM drives versus 10k RPM drives when cost is factored in.

SSD drives pay for themselves in power saving in about 5 years, well past their expected longevity.

Mechanical disks wear out more or less independant of their data load, SSDs wear out proportional to their data load.

SSDs do not handle tiny files very well due to how data is written.

I see nothing in the paper that is pro-microsoft, rather straight dealing on the drives themselves.

I would suggest MOD-TROLL any evanglest on any side of the OS wars as this paper doesn't seem to deal with OS touting.

It was a boring but informative read.

Re:Read the Paper (1)

OzPeter (195038) | more than 5 years ago | (#27507609)

Agree with you on your analysis. Just skimmed the paper as well and all those people who mentioned the paper writes pulling numbers out of their arse should go read it themselves.

Wished I had mod points for you

Re:Read the Paper (1, Interesting)

Anonymous Coward | more than 5 years ago | (#27507643)

One of the notable parts of the paper is they use 10k-100k write cycles as the maximum for the flash media(page 5, Reliability and fault-tolerance). Change that to the 1 million most manufactures are reporting and the numbers would be significantly different. Something else on this bit, is that the papers they reference for higher failure rates of DISK drives are 2-4 times higher so even using those numbers the failure rate microsoft used(10-100 times) seem to be skewed. And the authors appear to completely ignore this evidence for the evaluation of the disk drives.

Then change the prices to something more reasonable like $725 for 250gb, a fresh number from newegg.com for a OCZ ssd. Instead of the $730 for a 30gb drive and the numbers change radically again.

I couldn't find much information in English about the memoright drive(i spent maybe 15minutes on this total) they used but its appears on par with the spec for most other MLC based flash drives on the market currently.

I would guess they drive performs about 6-10 times better than their paper states.

Just another bit of FUD from m$.

Load More Comments
Slashdot Login

Need an Account?

Forgot your password?