It’s been said that airplanes magically turn money into noise.  In a similar way, it can be stated that disk storage turns money into heat.  In both cases, benefits like fast, reliable transportation and fast random data access accrue as the result of said magic.  Unfortunately, it also means a LOT of money may be involved yielding a lot of noise and heat respectively.

Setting aside the airplane analogy to concentrate on the conversion of money into heat through spinning disk over the long run, what can a storage administrator or CIO due to mitigate the cost of this magical transformation?  The answer to reducing the budget in various cases is tape storage.

Over the past four years, The Clipper Group has conducted a number of studies investigating the cost of disk and tape, including comparisons of power consumption.

The 2007 Clipper paper, “Tape and Disk Costs – What it Really Costs to Power the Devices” looked at a 5-year cost comparison between the power consumption of tape and that of SATA disk and concluded that, “The disk system costs over 25 times more money to power and cool than a similar tape system.”

In February of 2008, Clipper published, “Disk and Tape Square off Again – Tape Remains King of the Hill with LTO4”. In this case, Clipper investigated the 5-year cost of a tape backup system relative to that of a disk-to-disk backup system over the same period.  Clipper’s conclusion: “The energy cost ratio for a terabyte stored long-term on SATA disk versus LTO-4 is about 290:1.” 

Last December (2010), Clipper published their latest analysis, “In Search of the Long-Term Archiving Solution – Tape Delivers Significant TCO Over Disk.” In this instance, Clipper looked at a 12-year time horizon for both disk and tape systems and concluded the following, “The cost of energy alone for the average disk based solution exceeds the entire TCO of the average tape based solution.”  More specifically, “…disk consumes 238 times as much energy as tape under assumptions that lean toward favoring disk.”

With the average price per kWh consumed having increased by 33% over the past 10 years (Source: U.S. Energy Information Administration) holding data long-term is likely to become a relatively expensive energy proposition – especially if you try to retain all that data on disk.  If you project out the amount of data that may be stored, as the Enterprise Strategy Group has done (Spectra Blog: What could you buy for the cost to power an archive?), you can get a sense of what it might cost to power all that storage on either disk or tape.

The economics of tape energy consumption make it an ideal long-term storage repository.  Could that be Why Tape Rolls On?

To learn more about Why Tape Rolls On, see part one of this series discussing the Security characteristics of tape.

Between ESG and Clipper Group, one can generate an interesting discussion about the opportunity costs of powering archive storage on tape versus doing so on disk.  With the power savings available from tape archives, you could buy a lot of stuff.

For this exercise, let’s use ridiculously big numbers to see what we could buy with the money spent in power alone, just for archives.  The 2010 Digital Archive Market Forecast from ESG indicates there will be 79,151 PB of digital data archived by 2012 – next year.  By 2015, ESG expects that data mass to grow to 300,000 PB or 300,000,000 TB of data! 

The Clipper Group published its Clipper Notes in December 2010 providing insight into the difference in power costs between tape and disk.  The title of the article in question is, “In Search of the Long Term Archiving Solution – Tape Delivers Significant TCO Advantage over Disk”.  The article states that the cost of powering a disk archive is 238 times greater than that of tape.

Doing a wee bit of math with the Clipper data indicates the cost per TB of power stored on tape over the course of a year is roughly $1.04.  The Clipper folks calculated that the cost per TB stored on disk for a year is 238 times greater so let’s call it $246.70 per TB, just for grins.  Now the fun part.

If we multiply out the power cost per TB for both tape and disk relative to ESG’s archive forecast numbers for 2012, we get the following:

· Cost of power to store 79,151,000 TB of data on tape in 2012 = $82,317,040.  That’s a lot of beer.

· Cost of power to store 79,151,000 TB of data on disk in 2012 = $19,526,551,700.  That’s $19 followed by a “B”… as in billion.  Now we’re talking real money.  The kind only Uncle Sam spends.

Look further into the future at the power cost associated with archiving all the data ESG foresees and the digits become even more eye popping.  What’s more, these figures don’t factor in the escalating price of power over time.

· Cost of power to store 300,000,000 TB of data on tape in 2015 = $312,000,000.  You could buy a small fleet of yachts with that kind of coin.

· Cost of power to store 300,000,000 TB of data on disk in 2015.  Hold onto your hats folks… $74,010,000,000.  With a stash of cash like that you could buy the countries of Trinidad & Tobago ($26.4B GDP), Jamaica ($23.9B GDP), and the Bahamas ($8.9B GDP) for vacation purposes, then pick up a handful of other countries just to round out the collection – and still have money left over for that fleet of yachts.  Source: Wikipedia - so take the country GDP figures with a grain of salt.    

The point is this:  if you plan to archive large chunks of data for very long and do it only on disk, you’ll pay a lot of money for power.  If you elect instead to deploy an active archive using tape for much of that data, you can save a lot of money on power.  You may even save enough to pay for a vacation junket on your new yacht.

CTO Insight: Big Data: Why Tape?

 By Matt Starr, Spectra Logic’s CTO

I have watched the tape market’s growth over the last two years, which seems mostly due to the increasing number of archive installations.  With much larger system implementations projected through 2014, this growth will continue for the foreseeable future.  Military low-altitude and high-altitude video surveillance in countries like Afghanistan, the media and entertainment industry’s drive to 4K file data and the growth in PACS data are just a few of the many market segments driving the implementation of large archives. 

These are areas where dedupe and disk, in general, fall down, precisely because of the raw quantity of data involved--the disk resources required would be enormous, and use enormous quantities of power-- and the delays in time to deduplicate, then reduplicate is unacceptable.   

EMC’s recent “Big Data” news splash did not mention tape, which kind of shocked me!   (It’s only kind of shocking, as EMC is tape-hostile.) Tape is Big Data:  80% of the world’s data is stored on tape[1]and tape is the only media that can scale to exabyte(s) and still be cost effective.  In fact, tape is the only cost-effective method of storing Big Data.   Tape storage is denser than disk storage, costs less up-front and is ten times less expensive to operate over time than a disk-based solution.  I am not implying that disk does not have a play in the Big Data world; it is just not well suited as the “meat” of a storage environment.  

So, where does disk belong in this Big Data world?  First, disk works very well as the cache system that interacts directly with the user via a Filesystem, WebDAV, FTP or other front-end system.   Second, disk is the right platform for meta-data storage.  For far too long, users have been saving data as file names and not objects with meta-data.  As archives grow, object storage and meta-data will take the front seat in how data is stored.   Lastly, disk has an important role in helping to make stored data searchable: why would you store data if you cannot get it when you need it?   In my opinion, roughly 10% of the total archive space should be dedicated to meta-data and search.   Add another 10% of the total archive as disk space for cache, and the picture starts to come together.   Roughly 20% of your total archive should be disk, with the other 80% consisting of long lived, reliable, cost-effective tape.

Reliable? Yes. The facts are absolute and irrefutable-- tape is extremely reliable—more reliable than disk.  Tape’s error correction is 10 to -17^thup to 10 to the -19^thbits, which blows disk’s reliability[2]statistics out of the water.   Additionally, modern tape libraries have features like Spectra Logic’s Media Lifecycle Management that predictively informs the user about the health status of the tape as it being used. Features like this layer on reliability even beyond tape’s already high reliability.   Through MLM and other features (stay tuned for a few upcoming announcements this spring), Spectra’s T-Series libraries ensure that the data on the tape is intact and recoverable from the archive.

The architects and developers of data archives will continue to build systems based on disk and tape, not just disk.  When Big Data archives are based on disk alone, then one or more of the following scenarios is true:  1.) They are not a Big Data environment, but want to be (or think they are) 2.) They are wasting money and should be answering to their shareholders or voters.   3.) They have been mis-educated on tape.  In the end, tape is far from dead and will continue to prove itself as the ideal medium in the Big Data world.

by Mark Weis, VP of Spectra Logic Federal

Spectra recently joined General Dynamics EDGE® Innovation Network, a collaborative, open-environment initiative enabling industry and academia, with government input, to work together to enhance the delivery cycle of new technologies and innovative capabilities to warfighters and first responders. Currently, there are several projects in development at six EDGE Innovation centers worldwide. The projects are intended to improve and enhance warfighter capabilities.

Pete Palmer, EDGE Innovation Network director, said, “The EDGE bridges gaps between end-user needs and current capabilities by promoting rapidly prototyped solutions that could close those gaps quickly. By applying the EDGE process, the government can quickly and cost-effectively review multiple options for users to evaluate in months rather than years.”

So where does Spectra Logic fit in? Spectra Logic Federal delivers data storage and archiving solutions to more than 200 agencies across the USA. Our storage expertise helps agencies solve unique Federal-specific challenges. Spectra’s products are extremely customizable; products in production today. By joining the EDGE Innovation Network, Spectra Logic can utilize EDGE labs to further develop leading-edge technology to meet the specific data storage needs and requests of EDGE end-users by tailoring our flexible storage products based on their hands-on lab experience.

For innovative companies with federal expertise like Spectra Logic, joining the EDGE innovation Network keeps us in sync with what our customers want now, and allows us to add new features in our products to suit these needs. EDGE allows small businesses to display their products in a lab environment, a preferred option for many Federal prospects prior to implementation.

Stay tuned to the Spectra Backup and Recover blog for future Federal updates related to our new EDGE membership and other activities.

Dear Ms. Meade:

Did you read about the recent breakthrough[1]in tape technology—up to 50 TB per LTO tape? I was also “excited about the recently announced 3TB Seagate hard^[2] drive, but if tape devices can go to 50TB imagine what kind of hard drives we'll have in a few years time.”

Signed,

Seeing Only Disk

Dear Seeing Only Disk,

You might want to have your eyes checked. The breakthrough is in TAPE, not DISK. I find it so interesting that readers see the word “tape” and say the word “disk.” My, but the big advertising bucks spent by the Three-Letter-Disk vendor continue to pay off. Disk vendors have managed to obscure advancements in tape and instead bring to mind advancements in disk. Your disk dollars are hard at work with conflation in mind.

It is true that the 50 TB tape uses a perpendicular magnetic recording technology as does disk. In fact, tape and its automation are a lot like disk in other ways. For example, did you know that with Spectra tape libraries, you can invest in a global spare, much like you can have a stand-by in case of a RAID disk failure? And just as disk does several levels of data verification, did you know that tape drives perform a read-after-write and that Spectra libraries verify that data on tape can be retrieved through its PostScan™ feature?

So perhaps disk and tape parallels do hold true to some extent—except for the facts that

1. Tape doesn’t consume energy just to maintain data, as does disk
2. Tape media has an archival life of 30 years at a minimum, assuming decent conditions and the availability of a drive to read the data (not that big of a deal, if you store a few components along with the off-site tapes), while disk is typically used for five years or fewer; and
3. Tape costs less than disk, any way you look at it.

So aside from tape’s higher data density, longer shelf life, portability for disaster recovery, lower purchase price, ease-of-use through a file-system instead of data written in specific backup formats, and greater return on investment, disk and tape ARE a lot alike.

Or not.

Sincerely, and sincerely astounded at your reading of the letters T A P E as disk,

Ms. Meade E. Ahmogle

Who cares about a green data center?

Really?  Not in the negative sense as in, “Who the heck cares about a green data center?” but in the simple sense of, “Who is it exactly that cares about reducing power consumption in the data center?” 

This question was recently posed to me by a participant on a webinar I was leading.  The subject of the webinar was, “Why Tape?” and energy is always a part of answering that question.  But as for who would care about this, I didn’t have a simple answer.  The EPA did a study on power consumption in US data centers which was requested (mandated) by the US Congress.  I hate it when the US Congress starts asking me questions.  It’s never good…

The EPA study found that in 2006 data centers were paying around $4.5 Billion for power each year, and they projected that would grow to around $7.4 Billion a year by the end of 2011.  And that’s using “current efficiency trends.”  I quote. 

Who are these people?  While I don’t know exactly who would care about the power bill… at those prices… you’d sure as shooting think that somebody cares.

Is it time to delve into conspiracy theory?  Are the power companies behind this?  “Let’s keep them so confused about who cares what they spend on power that they’ll just keep on spending.”   As much as I love conspiracy theories, I think the answer is a bit more straight forward (aka boring).  Hope you’ll keep reading.

Historically IT teams have tackled evaluating and purchasing data center equipment while the data center facilities group has handled power and cooling.  The IT group usually reports to the CIO and the facilities group reports through facilities management.  In many companies, these are two entirely different organizations.  If IT is evaluating and purchasing the products that use the power and generate the heat, while Facilities is working to provide power and remove the heat, the guys buying the products (which suck up the energy) have no incentive to reduce the problem.  While I’m sure there are people who care about energy consumption in every company, they don’t care at the right time:  When the IT products are being evaluated and purchased.  

If you’re selling computer equipment, and you’re advertising power savings, you may be wasting your money.  If you buy computer equipment, and you ignore power savings as a feature, you may be wasting your power.   “Waste not; want not” my mom always said.  But she knew nothing of computers or data centers.  Yet I think she was onto something…

It’s not like this at all companies.  According to Dean Nelson, senior director of global data center services for eBay, eBay groups their power bill under the IT budget.  "When the CIO is paying the power bill, [he] really understands the impact of the decisions being made," Nelson reported at the Uptime Institute Symposium 2010 in New York a few months ago.  He went on to say eBay's IT department is "self-funding" new spending by being more efficient.  My mom would have loved this guy.   Although I am thankful her lack of computer skills kept her from becoming an eBay junkie. 

Do you know who pays the power bill at your data center?  If not ask around.   And if the person who is paying the power bill has no input on the products being purchased which use the power, ask “why”?  Your mother would be proud.

I recently had the honour of getting on stage at the Storage Awards 2010 to collect the gong for Green Storage Product of the Year, awarded to the T-Finity. The award win wasn’t the first for Spectra Logic over the last few months, but I’m not here to crow about our success (well, maybe just a little).

Instead, the award made me think about the whole Green IT movement and where we currently stand in the UK. The hype around Green IT has far outweighed the traction – in part because Green has been leapt upon by marketeers and the buzz-word brigade, but also because the ‘movement’ pretty much met the global economic troubles head-on as soon as it started to gain any momentum.

For vendors that started off down the Green IT messaging path there was a lot of backtracking as they focused their efforts on promoting cost-savings and TCO benefits in line with customers’ priorities. Now that economic recovery seems to be slowly creeping into view, there may be a lot of vendors coming full-circle, feeling that customers may be ready to move CSR up the IT agenda.

This posturing and positioning is what faces vendors trying to pass off any technology as IT’s answer to Captain Planet in a desperate attempt to gain a competitive edge. I’m referring to technologies such as vast disk-arrays that eat up moderately less power than a competing product while spinning away idly in the background. There is nothing Green about this – it is just a matter of being slightly less inefficient. Organisations mandated to be green will see through this spin, and instead hone in on genuinely Green technologies.

From Spectra Logic’s point of view, we like to think we can talk trees or £’s to customers – however you look at it we’ll save you both.  When we started designing the T-Finity did we set out to make the Greenest storage product on the market? No. But by creating the most efficient, scalable and dense tape library on the market we created an extremely green product by default.

So it is perhaps apt that commentators are referring to the green shoots of recovery – economic growth, after a period of such severe contraction, may well lead to a shot in the arm for Green IT. For our part, you won’t notice any difference in how we position our products – if you are a business that needs to Green your approach to IT – be it for legal or less self-serving reasons - then tape is the obvious choice for backup and archive.    

Dear Ms. Ahmogul,
IBM recently announced that it created a 35 TB tape that will be commercially available in five years. Since then I have read articles asking good questions, like ‘Do we need 35 TB tapes?’

Sincerely,
An Easy Mark

Dear Mr. Mark,
According to urban legend, in 1981 Bill Gates defended the (then-new) PC’s limited RAM by saying: "640K ought to be enough for anybody." (He denies having said this.)

This legend comes to mind on reading your question: “Do We Need 35 TB Tapes?” (Perhaps in thirty years you may find yourself denying having asked this question.)

To answer the 35 TB question: Why, yes, we will need 35 TB tapes. Just as we continue to require ever greater RAM than 640K, we will also need ever larger storage media, simply because of the incredible rate of data creation.

Simultaneously, the percentage of data to protect over the long-term is growing. (Please refer to the IDC chart, below.)  More data, with a greater percentage of it to be stored for a decade or more, leads to an unavoidable conclusion: we will need to store all that data someplace.

Naysayers revel in pointing out (the obvious fact) that it takes a long time to fill a 35 TB tape. Hence, these Captain Obvious types hold that no one will need or use a high-capacity tape. And, of course, it’s true that filling a 35 TB tape takes awhile. It also takes awhile to fill a 1 TB tape.

The fact that it takes awhile doesn’t alter other facts that include these:
• Huge quantities of data already occupy the digital universe and are multiplying at fantastic rates.
• A steadily higher percentage of this ever-increasing data has to be protected for the long-term.
• No viable alternative to tape for long-term, proven data storage yet exists (or is on the horizon, for that matter).

Over the coming decade, the choice is between using a few big tapes or lots of small ones.  So although you, Mr. Mark, may choose to use and manage 35 one-terabyte tapes, others will chose to use and manage a single tape.

The logic is pretty straightforward.

Kindest Regards,
Ms. Meade E. Ahmogul

Source: Gantz, John F. “The Diverse and Exploding Digital Universe,” IDC White Paper, 3/08.  “Figure 8 shows a unique view of the digital universe by the degree to which the information in it might be subject to […] legal and compliance requirements such as ediscovery, HIPAA, or Sarbanes-Oxley; or be valuable enough to expect to store for 10 years or more.”

Dear Ms. Meade:
How would an enormous up-front capital expenditure (aka CapEx) for a T-Finity reduce my capital expenditures? By definition, reducing CapEx means spending less, but you're saying if I fork out a lot of money, I'll spend less? Where is that logic? And reduce my operating expenditures? (OpEx)? T-Finity will do all that--is it a toaster and a floor wax, too?

Signed,
Doing Fine With My Powderhorn

Dear Doing Fine:
So glad to hear that things are going well in your world.

Would you be interested in what is going on in the real world?  It turns out that old technology is expensive. Tried to get spare parts for your Model T lately? How about finding truly floppy floppy-disks (the 8x8 inches model) for your Atari?

At some point, it's more expensive to keep old technology, given long-term expenses, than to replace it. Once we talk about replacing something, you are looking at your capital expenses. And the T-Finity does reduce Capex considerably. This part is very straightforward: costs less up-front, doesn't require extra software applications and  servers to run them--in fact, once you buy the T-Finity, you have what you need--partitioning (that's right, no database, no server and no external application, as required by the other very large libraries), encryption (that's right, no external software or hardware as required by the other very large libraries), and remote management (once again, no external anything needed). Right there, you've reduced your capital expenditures compared to the other guys.

Operating expenses falls right into line with this. Didn't I say NO external software, hardware, anything? Those typically  come laden with service agreements and idiosyncratic interfaces.  In my world, that means more parts to manage and to break, more software to learn, and more service agreements to pay. I’m not so sure about your world.

As to the floorwax part--T-Finity lets you use more floorwax, given that it uses so much less data center space. Alternately, you could put some other equipment in the saved space. And the amount of power a T-Finity uses is approximately that used by a toaster--but only when the library is really, really busy, and depending on how you configured it.

The expenses associated with your aging and soon-to-be no-longer-supoported Powderhorn are obvious and inescapable. Sometimes, you have to do the math and face hard financial realities. When that happens for you, please remember that the T-Finity is fabulous. (Yes, Ms. Meade E. Ahmogul gets paid by Spectra, but it's true anyway.)

I do hope you enjoy your Powderhorn and its savings on floor wax (as much as ten dollars). I also hope that, once you switch to T-Finity, you enjoy savings on space, power, and more (savings that will be in the thousands of dollars, even after you subtract the ten dollars on additional floor wax).

Sincerely yours,

Ms. Meade E. Ahmogul