Just wrote a piece for z/Journal about the two issues on the minds of every IT person in Florida these days — other than job permanency.  They are all talking about power and preparedness.  Power, in the summertime, is a source of woe from a cost, supply and reliability perspective.  Preparedness has to do with the inevitable power outages we will confront, whether as a function of everyone turning on their AC at once or Mother Nature wreaking havoc with us in the form of severe weather events.

After writing the column, I remembered a movie someone had given me that I recommend to everyone:  BombProof.  Ironically, the film, which was a good citizen gesture of Burroughs back in the day, stressed the need to protect data.  Their novel technology solution?  Microfilm.  Check out this link for more info.

I chuckle when I hear most disk array vendors arguing that tape is dead and disk is the best data protection modality going.  Sounds a lot like these guys talking about the permanency and protection afforded to data by microfilm.

Ironies abound.

I am developing a piece on DR requirements now.  Had an interview with ProStor yesterday regarding their InfiniVault product — the one with removable laptop SATA hard disks that they are claiming to be more cost-effective than tape.  Hmm.

Then it must be true. 

According to NetApp, quoting Cohasset, its de-dupe meets SEC regulations.  We can all sleep nights now.

Why the heck is Cohasset’s opinion any more credible than any de-dupe vendor’s?  The latter group has been making the claim for years that de-dupe is safe for data required to be kept in “a full and unaltered state” by the SEC and other regulators. 

We all have short memories, it seems.  This same consulting firm was paid by EMC to confirm that Centera was “compliance certified” awhile back.  My post on it is here — as are my exceptions to this finding.

Folks, if I were a disgruntled shareholder of, say, Apple Computers, put off by the fact that Steve got the liver that my kid was supposed to get because he was much richer than my kid, and I decided to file a writ or warrant for some SEC data from Steve-O’s storage, my smart attorney would do everything he could to beat up the Apple-heads, including casting aspersions on the integrity of the data that was provided on the basis of its previously de-duplicated state. 

It wouldn’t matter one wit whether the de-dupe technique used did not IN FACT alter the data, only that it MIGHT alter the data.  The resulting case would probably go on for years.

The meaninglessness of Cohasset’s findings, bought and paid for as they are, is profound.  Are they offering to pay my attorney fees?  Is NetApp?

From where I’m standing, we are still in legal limbo in a litigious world when it comes to de-dupe.

EMC is standing by its offer to Data Domain shareholders while NetApp continues to press its courtship for cash and stocks.  Kind of reminds me of this scene from an old favorite movie…

Let ‘em duke it out.  We’ve got bigger fish to fry than speculating on the outcome of the acquisition of a de-duplication in hardware play.  De-dupe, as CA and others have demostrated, is a feature — not a product.  Data Domain did a good job of establishing the viability of de-dupe, but selling overpriced disk and overpriced server “head” to support what is essentially a software component is ultimately a fool’s errand.

What is important to me is a conversation I had today with Don Kleinschnitz, who is now a Senior VP of Engineering at CA.  Don and I go back more years than we care to admit.  He recently transitioned from Symantec, where he should have been in charge of the whole storage business unit after that company bought Veritas.  I wrote words to that effect in an article at the time, creating a bit of a political correctness issue as I recall.  But, heck, had Don been put in charge of the Veritas stragety, Symantec’s offerings in storage management might have been much more significant.

From today’s chat, Don and I remain fairly closely aligned in our thinking.  We aren’t exactly in sync, we won’t be taking any warm showers together anytime soon, but we agree on most things.

We agree, for example, that W3C Web Services standards might actually be the key to gluing together applications and infrastructure — to provision to data the right services (resources and functions) per policy and without much hassle.  To his credit, he has a lot of web services-savvy developers working on the next iteration of CA’s data protection products. 

What that means to me is that applications or policy engines will shortly be able to call up an appropriate data protection service — backup, CDP, HA — to provision to certain data based on the criticality of the data.  Heck, if he takes it to extremes, you will also be able to map certain data to a disk VTL target and expose it (or exclude it from) de-dupe whilst it rests on that target for, say, 30 days. 

Don clearly gets this concept, which I think we discussed — sans web services standards — a decade or so ago.

Where we part company a bit is in Don’s embrace of server virtualization and the whole cloud phenom.  I agree that virtualization is inevitable.  I just don’t think that all of the necessary tools have been refined to a point where I would trust them with my data.  As for clouds, I see them as a standards-free zone currently.  Nobody’s cloud is going to talk to anyone else’s cloud without a lot of jury rigging.  Again, web services could be a boon to cloud interoperability, but I don’t hear too many vendors of primordial clouds talking the talk or walking the walk of open standards.

Don and I started a bit of a debate that will probably go on as a blog war when it came to the subject of the future of the data center.  I said something like, “Cloud rhetoric is framed so as to scare IT operatives with the prospect that their days of employment are numbered.  It is hard to sell a technology idea that many prospective buyers perceive as putting them out of a job.”

Don countered with something like, “Jon, you ignoramus.  IT is under the gun for being terribly inefficient and terribly costly.  Clouds will replace data centers as we know them, whether or not IT likes it.”

This is not a fully accurate account of the conversation, inasmuch as Don does not talk like a Brit.  However, the gist is that we agreed that the more commoditized hardware becomes, and the more it avails itself of real virtualization, the more that IT can shift its attention to the things that matter:  not managing spinning rust, wires and chips (as in Information Technologist), but managing data and applications against the backdrop of business requirements (what Don and I once called “data processing” in our mainframe youth).

Still, Don is more optimistic than am I that clouds will amount to something.  That should make for an interesting dialog between our blogs.  Stay tuned.

3PAR has its panties in a twist about the proposed Energy Star standard for storage.  You can read about the standard and about 3PAR and others reactions to it here.

The quote that jumped out at me from the rant is this one.  Rather than focusing on energy efficient power supplies, the writer argues the following.

There are far bigger savings to be made with software functionality in storage arrays. Software allows two main benefits to be achieved:

• Storage array software can reduce the power or turn off disk drives; disk drives account for 80% of the power of a storage array.
• Storage array software can reduce the amount of data that is stored on the disk. Techniques for reducing data on the disk include:
  Virtualization
  Thin Provisioning
  Deleted data reclamation
  De-duplication
  Compression
  Tiered storage
  These techniques are in the main additive. The potential for saving data using one or more of these techniques is again to achieve an additional 80% or greater reduction in the amount of data stored.

Hmm.  Apparently, the Energy Star guys are a lot smarter than I gave them credit for.  By ignoring a lot of value add software embedded on array controllers, and the dubious value of this stuff to actual energy use reduction, perhaps they recognized the real issue.  Saving power in storage doesn’t derive from compresssing and de-duplicating bits in a thinly provisioned and tiered infrastructure.   Those are tactical approaches that deliver short term gains but no real long term improvements.  The only way to reduce power consumption by storage is to reduce the amount of disk storage.

Yup.  And the only way to reduce the amount of disk storage is to manage data better.  That means getting rid of the dupes and drek, the contraband and the archive data that currenly occupies about 2/3rds of every spindle you have.  Green archive on tape or optical (or even removable HDs like Infinivault) trumps disk storage anyway you cut it.

That 80% savings they are on about from doing value-add on the array controller makes no sense over time.

There, I said it.

Curtis Preston has done his best to cobble together some charts of the claims of dedupe vendors.  If you are considering this technology, it is worth a read.

The data is from spec sheets and published vendor numbers, not from independent tests.  To his credit, Mr. Preston acknowledges this and supplies all of the necessary caveats.

Good job, Mr. Preston.  This is useful data because of its aggregation in one place.

Folks, I’ve got the blues today.  Didn’t mean to get them, ‘just did.

I spent an hour this morning on the phone with a storage start-up in Minnesota, Pranah.  Here is their press release.

 Pranah Set To Redefine Data Storage
for Small and Midsized Businesses

Combining Highly Efficient Hardware and Software to Produce
the Most Complete Network Storage System Available in a 1U Footprint

SAINT PAUL, MN (PRWEB), JUNE 23, 2009  –  Pranah Storage Technologies (www.pranah.com) today announced it has created the easiest-to-use and most complete network storage system for the under-$50K price class.  (This class is the fastest-growing part of the storage business, according to IDC’s latest report on storage industry revenues, issued June 5.)  Pranah’s system features purpose-built hardware, designed specifically for data storage, combined with storage software applications that are typically seen only in high-end systems.  Designed for simplicity and scalability, the company says its Pranah 2000 Series provides more value than any storage system available to small and midsized businesses (SMBs).

“We’re doing nothing less than redefining reliability, flexibility, and value in storage networking for this underserved market,” said Steve Carter, Pranah’s CEO.  The company’s intelligent network storage products incorporate both standards-based and patented proprietary technology.  By developing and controlling its own hardware and software, Pranah improves ease-of-use for sophisticated storage applications. “In a difficult economic environment, our solution allows customers to cut their costs both for the amount of hardware required and the staffing associated with overly-complex systems,” he said.

Performance and Scalability
Pranah’s 2000 Series storage controllers can be aggregated, thereby delivering increased storage performance previously unavailable on systems in this price class.  The company’s support of 10Gb/s iSCSI allows it to stand out from the competition and delivers unique advantages in supporting applications requiring massive storage demands, such as video surveillance/security, medical imaging, document retention, video production, and cloud computing.

Power, Cooling, and Density Improvements
The new product line features “green-friendly” design that utilizes a U.S.-patented side-vented rail system to improve cooling by 31%, and allow a storage density improvement of up to 50%.  Pranah 2000 Series systems do not require dedicated data racks, and therefore can share the same rack with IP switch and routing equipment, telco equipment, and racked servers.

Product Details
The elegant design of Pranah 2000 Series systems is based on simplicity, delivering more value in a 1U form factor than any other product on the market today, including more disks per U, and featuring plug-and-play expansion with SAS, SATA, or Fibre Channel drives.

• Each system comes with standard software (called the Pranah Control Suite™), which includes Thin Provisioning, Tiered Storage, Snapshot, and Replication, all managed from a simple browser interface.
• It features a Distributed File System (DFS) that enables geographic diversity of critical data.
• It integrates both SAN and NAS within the same (1U) storage unit. 
• It can include multiple storage controllers and multiple host connections.
• It features a patented cooling system.
• And it’s scalable without extensive product support or costly reinvestment.

Value Proposition and Differentiation
Because Pranah has developed its own hardware and software intellectual property, it offers a unique value proposition to its customers.  Elegant design and tight integration make the system simple to operate and easy to maintain.  A key benefit of the company’s systems is that NAS is integrated as a standard feature.  By doing so, the customer does not have to do its own time-consuming, expensive integration, and maintain third-party software and hardware to support file-based data.  Storage applications such as Thin Provisioning, Tiered Storage, Snapshot, Replication, and Distributed File System (DFS) were developed by Pranah to be easy to use yet fully functional, and they all come loaded on the system, easily activated with a software key when required.

Company Strengths
Pranah has a seasoned management team, with a proven history in the data storage, software, and military markets.  Its development team has a combined 200 years of hardware and software experience in bringing storage and server products to market, and a strong intellectual property position.  Pranah’s products redefine storage networking by simplifying operation, increasing reliability, reducing complexity, and significantly lowering both cap ex and op ex costs.

Product Availability
Pranah’s 2000 Series product line will initially include three different control units, and three expansion units.  Product is shipping in late July in limited availability to 15 key customers and resellers for testing and evaluation, with general availability in September.

Advantages for Smaller Businesses
“Now, businesses that balked at six-figure price tags for a shared storage system have an better answer,” said Dave Walstad, Pranah’s VP of sales and marketing.  “Pranah’s systems were designed with the small business customer in mind.  To gain further leverage on a Pranah investment, Disaster Recovery can be easily deployed, since shared storage can be easily replicated off-site.  What’s more, SAN and NAS are delivered in the same box, so there’s no need for two separate systems, with the attendant complexities of integration and maintenance.  We’re confident ours are the lowest cost, smallest footprint, easiest-to-use storage systems in the under-$50K price class.  And they have the flexibility to meet the ever-changing needs of growing businesses.”

These guys want to compete with HP/LeftHand and Dell/EqualLogic as the gear provider to the SMB.  They say that they have a lower price than Compellent and also have a NAS head, which Compellent really doesn’t deliver without a lot of pain.  $10K gets you a single controller, an iSCSI connection and 4 SATA 1TB drives.  They have some cool (pun) patented technology for side venting 1u rack mounted storage, and they have no screws or cables connecting drives to the housing, but instead borrow some military specification shock resistent mounting foo.  Like 3PAR and EMC, you can aggregate their controllers to support better performance and more spindles.  Thin provisioning allows you to add more drives anytime.

You can source drives from your reseller, but if you do, the reseller becomes your tech support provider if drives fail.  Pranah will stand by its controller.

They said that they were compelled to embedd all of that “value add” technology on the controller, rather than going the Xiotech route, by financial wiz kids (VCs) who told them that, since the big boys were doing it, they should too.  I tried really hard not to cry, but it gave me the blues.

Then, I had another call, which I had forgotten about, with a San Francisco start-up offering a data protection service that leverages Data Domain boxes.  If you don’t have a DD appliance to mash your data so it can be transferred to a DD appliance at the data center where they have some cage space, you need to buy one.  They tout “the only SLA in the evaulting business” and say that their service and price point is targeted to the sensibilities of the midsized company that doesn’t like tape and doesn’t really want to do anything about classifying data assets to determine which ones require protection.

The smart fellow I spoke with said that his service simply reflects the prevailing sentiment in the market.  “Tape sucks” — because no one really knows how to do a backup.  “Tape is dead” — because Gartner was paid by EMC to say so.  “De-duplication is hot.”  Customers don’t really want to sort the junk drawer and encouraging them to do so — even if it is for their own good or would improve the efficacy of their continuity strategy – ”lengthens the sales cycle.”

Hard to shake, them blues.

Oracle has decided to kill Virtual Iron after buying the company only a scant few weeks ago.  The move leaves several thousand customers holding the bag with 500 to 1000 licenses each according to The Register.

Why did they buy VI?  Were they just set on taking the company out of their sky?  Acknowledging that VI was a small fish in a small pond, they wouldn’t have posed much of a threat.  Why couldn’t Oracle simply have licensed the technology they wanted (Walsh probably would have been agreeable) rather than buy the entire company and shaft its customer base?

Hmmm.

I was surveying leading storage vendor announcements today on their websites after following a flash animation link on InformationWeek pointing to a NetApp claim that they were going to significantly reduce our collective spending on storage. Seems like everyone is adding unheard of amounts of “value add” technology to their arrays. Technology to reduce data bit descriptions, to dynamically reallocate storage capacity once given to an application, to move data from one expensive set of spindles to another overpriced set of spindles, to protect data through unique bit writing schemes or mirrors, to create mirror split image copies for crash coherency, etc. The messages are all tuned to the current Do More With Less (as long as you buy only our stuff) channel.

I wondered how they were supposedly realizing all of their purported performance gains while adding so many latency inducing functions into the controller. Are they spoofing? Are they enlarging memory caches? How do they defy the laws of physics?

Then I found this hidden picture in a secure archive showing the insides of an IBMEMCHDS controller:

Of course!  They are using undocumented aliens to manage their many I/O headache-inducing technologies.  Just stick the little fellows into the array controller, and so long as you don’t feed them after midnight or get them wet or expose them to sunlight, all of your storage problems magically disappear.

I say, stop the oppression!  Free the aliens in the controllers!

A fund is being set up as we speak. 

By the way, where’s PETA when you need them?  Too preoccupied with fly swatting presidents than the tragedy of aliens held hostage in storage controllers.

I just received this email announcing a webcast featuring ESG and Isilon Systems claiming to address the key issues of storage today:

Storage issues are mounting, and so are the critical business and technology decisions you need to make. There are many “issues of the day” that you should be following closely; here are just a few:

• Adapting storage to virtualized compute resources
• Creating a scale-out infrastructure to support growing data capacity needs
• Lowering storage operating expenses
• Deploying data de-duplication capabilities
• Reducing costs with a tiered storage approach
• Archiving to disk versus tape

You need more than raw data: you need insight and perspective, and the opportunity to ask your business-critical questions. The Enterprise Strategy Group (ESG), Isilon Systems, and Ziff-Davis Enterprise have teamed up in this one hour free webinar to provide you the latest market research insights and an open forum to raise your toughest storage questions with Isilon’s scale-out NAS experts.

Come prepared with your toughest storage questions: the answers will be here. Register and attend, and you’ll be eligible to win a Kodak HD video camera.

I am wondering what readers here think about the priority assigned to the issues above.  Are they really key?  My view:

• Adapting storage to virtualized compute resources:  If you are doing VMware, Hyper-V, etc., this is probably an issue of importance.  I don’t know what, if anything, can be said about it currently other than to deploy Virtual Instruments to get an honest read on I/O so you can route it more effectively.  Of greater importance is understanding what the I/O is so that intelligent decisions can be made about the resources and services that should be availed to it.  That requires an understanding of the business context around the applications that are being used.  Plumbing issues are secondary.
• Creating a scale-out infrastructure to support growing data capacity needs:  Prioritizing scale-out is appropriate assuming 1) that you have done everything you can to take junk data off spindles and have implemented a green archive (tape or optical) to store non re-referenced data that needs to be retained none the less for business reasons, and 2) that you have done your best to classify data assets so they can be managed effectively over their useful life.  If you haven’t done either of these things, you are doing nothing but playing into storage vendor hands by working to build more capacity to store junk.
• Lowering storage operating expenses:  These days, cost-containment is job 1.  You lower expenses by eliminating stovepipe arrays and by establishing a common management scheme for storage (perhaps a specific management software package or a web services-based approach) and telling your vendor that you won’t buy their overpriced spindles unless they conform to the management approach you have selected.  That would really help wrangle in infrastructure management costs so that fewer storage admins can do more with fewer hands.  Again, intelligent archive and better data hygeine will help trim the junk drawer so capacity management isn’t a game of putting out daily fires.  In other words, operating costs are a function of common platform management and data management, not throwing more crappy value-add software into an array controller.
• Deploying de-duplication capabilities:  This is certainly a technology on a lot of peoples’ minds.  I will be doing a webcast on de-dupe AT THE EXACT SAME TIME (see below) as the ESG/Isilon cast.  Mine dwells on the business ramifications of de-dupe, whether the technology constitutes a product or just a feature, whether it is best done in software or in hardware, and whether we need to worry about the impact of de-dupe on things like regulatory or legal acceptability of data.  I suspect that a hardware vendor might simply encourage the application of value add de-dupe technology on his array controller.  What do you think?  Perhaps you can listen to one webcast or the other then listen to the playback of the other cast the next day to see which one addresses your concerns about this issue.
• Reducing costs with a tiered storage approach:  I am hearing vendors elevate this to an issue all over the place.  On array tiering is being manufactured as a cutting edge new feature of brand X arrays.  I have yet to find a consumer who is really buying into it.  Putting shelves of FC/SAS and LCFC/SATA into the same array is not intelligent tiering — it is just a way to jack up the price of the disk drives.  Migrating data from shelf to shelf based on simplistic watermarking or date last accessed+FIFO algorithms is not intelligent tiering — it is simplistic HSM.  I could use Novell Storage Manager, Crossroads Systems File Migrator, QStar Technologies wares, or Digital Reef to realize the same functionality (or more granular data movements based on data class) in a much less expensive way across Xiotech ISE platforms or generic JBODs.  Tiered storage is fine, as long as you recognize that unlike the mainframe, there are only two tiers of storage in open systems:  capture storage (rated to the speeds and feeds of the app writing the data), and retention storage (everything else).  Archival data belongs on tape or optical.  Junk data belongs in the waste basket.
• Archiving to disk versus tape:  Bizarre that we are even discussing this.  Have we collectively forgotten everything we ever learned about the vulnerability of disk, its inappropriate and costly application to long term archive, the energy costs associated with spindles versus other media?  If ESG is recommending archive to disk, I have to wonder about the intelligence of some of its other recommendations.  (Well, I guess I do anyway…)

Okay.  I’ve said what I think.  What are the real issues on everyone’s mind when it comes to storage?  How about these?

• When is my management going to let me purpose-build my storage infrastructure to meet the needs of my applications and business processes?  When will they realize that one size fits most products of brand name arrays don’t fit anyone’s needs very well.
• When is the industry going to cooperate in a universal storage management scheme that makes it easy to unplug vendor A’s wares and replace them with vendor B’s?  Better yet, when will the industry stop preventing efficient management by adding a lot of crap to their array controllers that doesn’t need to be done on the array and probably shouldn’t be?
• When is the industry going to quit equating storage management with capacity management?  Management of storage is about providing the appropriate hosting services – conceived as a well defined and highly manageable mix of hardware resources and software functions — to data over its useful life.  It isn’t just waste management:  creating an ever expanding landfill in which to deposit junk data.
• Why scale out;  why not scale back?
• How about abandoning terms like SAN, networked storage, tiered storage, de-duplication, archive, etc. that have become so much abused by marketeers that they mean whatever the vendor wants them to mean in his brochure?  Instead, let’s talk about what really matters:  I/O performance, data management via classification and routing policies, driving costs out of boxes of commodity spindles, using the right storage technology for the job instead of what the vendor wants us to use…

Okay.  Back to earth.  Competing with the ESG/Isilon webcast, at exactly the same hour, is my webcast for Redmond Magazine on De-Dupe.  Here are the details as they were just sent to me.

De-duplication is a hot topic in storage management, but its primary application is typically in the realm of data protection. Jon Toigo offers his thoughts on the de-duplication craze and notes a great option that will enable you to leverage de-duplication capabilities in a sensible and business-savvy way. Join us for this free webcast!

• Date: Wednesday, June 24, 2009 at 2:00pm ET
• Webcast: The De-duplicated Backup: Straight Talk about the Capabilities and Limitations of a Sexy New Technology
• Speaker: Jon Toigo, Enterprise Systems Contributing Editor

The idea is to write backup data sets to an array or virtual tape library, then to apply an algorithm that squeezes the number of bits used to describe the data. In theory, this lets you build a dense backup file repository – storing more data in less space – from which individual backup files can be recovered quickly in the event of accidental erasure of, or damage to, originals.

The question is, how do you best deploy de-duplication technology? Hardware vendors want to sell an appliance or gateway. However, these products can be prohibitively expensive to purchase and costly to operate.

Does the strategy eliminate tape backup, as some vendors claim? Are there any potential compliance issues associated with the strategy that should lead you to exclude certain data from de-duplication?

Registration is here, or you can watch the ESG/Isilon cast and visit the replay of my webcast the next day.  Or you can just go to the beach and enjoy some Summertime with your kids. 

Your choice.

UPDATE —

I am now advised that the ESG/Isilon cast is at 4PM EST (I misspoke using the PST time).  So, you can blow off your workday and join both casts, or blow off the webcasts and hang out with your kids, who are likely bouncing off the walls given that there is no money for special camps and staycations have replaced vacations in many homes.

I’m not sure about all of the hardware stuff, but I think HDS is showing a lot of smarts about contextualizing their storage wares within the realities confronting IT and business today.  Here is an extract from a press release they sent me on Wednesday.

Hitachi Data Systems Introduces Breakthroughs in Storage Cost Savings and Dynamic Provisioning
New Software and Services Help Customers Leverage Current Investments to Reduce Costs, Increase Flexibility and Optimize Performance

SANTA CLARA, Calif. — June 17, 2009 — Hitachi Data Systems Corporation, a wholly owned subsidiary of Hitachi, Ltd. (NYSE: HIT) and the only provider of Services Oriented Storage Solutions, today introduced unique capabilities across software and services to enable customers to reclaim underutilized storage capacity and increase the return on their assets. This announcement highlights the company’s strategic Global Services capabilities to further extend the economic and optimization benefits customers can achieve leveraging Hitachi Dynamic Provisioning technology.

[Okay, so I disagree with the hype around being the ONLY provider of Services Oriented Storage Solutions.  I would have to give that kudo to Xiotech, but...]

Coupled with the recently announced Switch IT On program, the enhancements can help customers improve the efficiency of their existing environments by reclaiming storage capacity across all storage tiers. Customers have realized up to 200 TB in reclaimed storage capacity and CAPEX savings upwards of $2M, thereby deferring future storage purchases.

[200TB equals $2M?  Think of the mark-up inherent in this equation. Obscene, isn't it.  We would be talking $40K for 200TB of SATA 1TB disks raw, or about $240K for 200TB of Seagate 15K drives...  I guess the other costs are "soft costs"?]

Specifically, Hitachi Data Systems today announced the following innovative capabilities:

• Zero Page Reclaim – examines volumes of physical capacity on or on storage connected to the Hitachi Universal Storage Platform®, returns unused storage blocks back to the storage pool and reclaims storage space
• Industry First Enterprise-Class Automatic Dynamic Rebalancing – when physical volumes are added to expand a dynamic provisioning pool of storage, existing virtual volumes in the pool are automatically re-striped across these new physical volumes to rebalance the workload. Hitachi Data Systems is the only enterprise-class storage vendor that automatically rebalances the pages of a virtual volume so they are actively re-striped to take advantage of new disks when the pool is expanded.
• Support for the Hitachi Adaptable Modular Storage 2000 family – brings the advantages of dynamic provisioning such as cost savings, automated performance optimization and easy provisioning, to Hitachi midrange customers
• Storage Reclamation Service – assesses the customer’s environment, plans the new dynamic provisioned environment, migrates the data and reclaims unused capacity with Zero Page Reclaim without disruption to the application, thereby deferring CAPEX and increasing ROA.

“Storage economics, our best practices for helping companies assess their storage requirements and investment returns through true ownership cost, has never been more critical to enterprises and thin provisioning can play an integral role in saving organizations money while improving operational efficiency,” said Hu Yoshida, Chief Technology Officer, Hitachi Data Systems. “Today’s software and services announcement is part of our focus to help our customers do more with less. Dynamic provisioning simplifies operations by replacing the management of hundreds of volumes with the management of one or two pools of virtual capacity. Dynamic provisioning can also reclaim capacity on existing open systems volumes without disruption for significant cost savings that go straight to customers’ bottom line.”

[Gotta love Hu.  But why do dynamic provisioning on an array controller when I could just use DataCore or FalconStor and much less expensive disk?  An honest question...]

New Hitachi Storage Reclamation Service
Hitachi Data Systems has extensive experience helping customers take advantage of its dynamic provisioning technologies to increase utilization, reduce the cost of storage growth and extend the life of storage assets. Through the Storage Reclamation Service, customers can gain the benefit of this experience to accelerate time to results with lower risk. Key aspects of the service include:

• End-to-End Service – includes assessment, planning, design, migration and storage reclamation 
• Quick Analysis – to show the type of benefits that customers can expect, Hitachi Data Systems can perform a quick analysis of a sampling of a customer’s environment
• Shared Risk – to lower the risk to customers*, Hitachi Data Systems will perform the assessment in a shared risk model, where customers pay only if Hitachi Data Systems exceeds the pre-set customer expectations for storage to be reclaimed.

Interesting press release, but I like the lower half better.  I think everyone needs a storage assessment that can help them reclaim all of the expensive spindle space they populate with archival data, orphan data, and contraband data today — and so they can spot capacity holdbacks that most leading array vendors routinely take from the capacity they sell.  Also, why don’t we fix pricing based on customer-usable capacity rather than raw unformatted capacity.  Hey, Hu, get the bosses at HDS to follow Xiotech’s lead in this space…

When you do the math, it sounds like some overpriced storage is being sold together with a paygo service for reducing the amount of overpriced storage you need to buy.  I am not singling out HDS for any criticism, of course.  This is an industry wide phenom…excluding Xiotech.