We will shortly have our second SAS array in the lab and hope that it performs better than the one we already have. As previously feared, there are microcode issues with some of the early SAS controllers we have been seeing. Apparently, the presence of a standard has not dissuaded some vendors from adding proprietary foo to their equipment.

I keep getting pinged by folks regarding the differences between SATA and SAS. Is SATA full duplex? Is it enterprise-ready (whatever that means)? Do the drives really shake so much that over the course of the evening the drives will shake themselves out of the array cabinet and end up in a pile on the floor? The answers are, in order, no, who knows?, and bullshit.

A good debate between Fujitsu and Western Digital on the relative merits of the technologies can be found here.

End of line.

Tony Asaro’s blog at Computerworld was just sent to me today by a well-wisher. In 750 words or less, he tried to explain why there are so many storage vendors and how they differ from each other. Wow! That’s a huge challenge, Tony.

He focused on EMC and Network Appliance. He said EMC was offering a multiplicity of platforms, compared to NetApp’s one-platform-with-different-software-fits-most strategy. That was a key differentiator in his mind. He also said that NetApp has a vision that they are rolling out slowly, under no pressure since sales are good, while EMC lacks any vision as far as he could see.

The reason for all the other platforms out there was that everyone had a slightly different take on CAS, ILM, HSM, data compression, etc.

So, there you have it. There is Network Appliance, EMC and All the Rest.

I have to wonder sometime about who is writing about storage. I don’t know Mr. Asaro, but it seems that this analysis reflects a bit of naivete.

EMC, in my humble opinion, isn’t one company. It is many companies. Simplisticly, they are at least a bifurcated company, split along “party lines” — selling proprietary iron with integrated software to lock in consumers, on the one hand; struggling to migrate to software and services and to get out of the box business (or at least to minimize dependency on increasingly commoditized platforms) on the other.

Network Appliance is pursuing a course that EMC claims to have already tried and abandoned: “You can have that box in any color you want, as long as its black.” The fact that NetApp can sell a lot of commodity boxes (Warmenhoven’s words) with proprietary parts (Linett’s words in a recent blog) and call the result “open commodity architecture” — an oxymoron in my view and NetApp CTO emeritus Dave Hitz’s (read his recent blog) — only goes to the point that a former EMC CEO made a couple of years ago: We are all selling a box of Seagate hard disks.

While it is compelling to focus on the Billion Dollar Babies of the industry, the really groundbreaking stuff is happening in the also rans. The incumbents must meet the needs of a huge install base, allowing for little innovation compared to the newbies who have no install base to protect and are thus able to think outside the box.

When you look at it, everybody, from large to small, sources their drives from the same core cadre of drive manufacturers. They get their frames from the same core cadre of frame manufacturers. Their IP is in controllers, backplane designs to some extent, and software value-adds. Recently, a lot of money has been made in interconnects and interfaces for switched direct attached storage (FC SANs and iSCSI).

A lot of the storage plays out there today are not really about boxes at all. Their investors have compelled them to wrap their software in tin-and-spinny-things just to have more to show for their money than a PowerPoint deck.

Financial guys are looking for multiples. What technologies will become hot as current technology becomes same-old same-old? The game plan these days is to watch the field of little guys, see if anything attracts enough early adopters to qualify for big league attention.

The situation is rather like a favorite old movie of mine, The Blood of Heroes, which follows a futuristic society in which the poor live above ground in a post-apocalyptic world where life is nasty brutish and short, and the richies live in underground cities, presumably where they have resided since the apocalypse itself. Their one common thread is a kind of 23rd Century version of football played by five or six guys per team.

On the surface, there are the minor league teams who wander from village to village, competing in a vicious sport whose objective it is to place a small animal skull on a post. When you have won enough games and collected enough skulls, you take your team down into the underground cities and present all of your skulls in the hopes of having an opportunity to compete with a big league team. The objective isn’t to win, it is merely to play well enough that some of your team will be recruited by the majors.

In the story of the film, our gaggle of heroic players not only competes, they actually beat the major league team they are playing. Huge event. Unprecedented. Heroic. Secretly, that is the hope of many little storage companies out there. Only, their venture capitalists and investment bankers often do not want them to win. Nobody wants to IPO in the current market. They just want to get the attention of the big leagues and sell the IP off — often into oblivion.

Just a ramble. Good movie though.

In my next column on Byte & Switch, I am featuring an interview with Paul Carpentier, the fellow who created FilePool technology, which was ultimately purchased and wedded to a proprietary controller by EMC to create its infamous Centera platform. Carpentier and I have been going back and forth for some time, meeting at conferences and exchanging emails about the dilemma of current hardware CAS solutions. As the guy who officially thunk this stuff up, he knows where the bodies are buried — especially in the EMC implementation of his wares.

I liked what he had to say and ran much of it intact in the Byte & Switch piece and, assuming that editor Terry Sweeny likes it too, you can read it in its full glory pretty soon. What I left out of the piece was some of the conversation about CAStor, the product that Carpentier’s company Caringo will be introducing shortly. I didn’t want the column to sound like a product plug.

For those who are interested, here is a little more information about Carpentier and his burgeoning new technology.

Give me background on Paul Carpentier and on Caringo (if you can talk about Caringo yet) as it pertains to CAS.

• Paul Carpentier has over 25 years of experience and has served as founder, CEO and CTO of four successful technology companies. He holds six fundamental patents in various computing technologies, and holds several graduate and post-graduate engineering and business degrees. Four of Paul’s patents were acquired by EMC and another two are the basis for the next generation of CAS that he is shaping at Caringo.
• Mr. Carpentier founded Wave Research in 1992 where work on software distribution and management systems led to the development of FilePool, the fundamental technology at the origin of CAS. EMC acquired the FilePool company and technology in 2001, and used it as the foundation for the extremely successful Centera product line.
• Confident that this was only a beginning, Mr. Carpentier continues his tradition of innovation at Caringo, Inc., which he co-founded and where he serves as chief technology officer. His patent-pending upgradeable and scaleable information integrity model will enable a fundamental shift in the way reference information (fixed content) is stored, secured, guaranteed, audited and managed.
• Paul co-founded Caringo in early 2005 with 2 software veterans. Jonathan Ring, Mark Goros and Paul have over 80 years experience between them and span over 15 startup companies. The team first met 15 years ago and has worked together in a number of companies since then.
• Caringo is leveraging an international virtual company infrastructure where an extreme programming model is used to bring out advanced production software products in a compressed timeframe with a small, focused team located in Austin, Texas.

Lay out key differentiators between the Caringo approach and other hardware-centric approaches to CAS.

• CAStor is a software product and runs on commodity hardware. It provides a fully integrated, bootable operating-environment with zero installation and configuration requirements for per-node deployment times that remain under five minutes.
• CAStor was architected from the ground up for high performance. We anticipate customers using CAStor for much more than merely archiving fixed content.
• Dead disk or node recovery time is inversely proportional to the size of the cluster – the more nodes are available, the faster it goes.
• CAStor will automatically absorb a new node brought onto the network and will scale seamlessly to beyond 500 nodes in its first iteration, with 5000 a very attainable mid term target. Each additional node will increase the cluster’s overall throughput.
• An industry breakthrough, patented transparently upgradeable hash is part of the Pollux Content Integrity Assurance option to CAStor.
• Interface to the CAStor is via industry standard HTTP. We are also looking at being among the first to offer XAM support – if the initiative catches on with customers.
• Caringo believes in science, measuring and numbers. As such it rejects the current policies of many vendors in the CAS space that not only do not publish actual usable performance numbers, but even contractually forbid their customers from publicly providing test or benchmark results. Caringo will work with any party wanting to test and measure relevant CAS performance numbers and make them available to the public; it invites all market participants to join them in this necessary effort to bring the CAS market to the next level.
• CAStor’s radically simple per-disk pricing model offers solutions for nearly an order of magnitude below Centera and others.
• Combining the customer’s choice of commodity hardware with low pricing and zero administration, Caringo will be hard to beat in providing the lowest TCO in the fixed content storage space.

What’s your current status and timetable for delivery?

• Caringo is currently delivering CAStor beta to a few partners. Limited Availability with full functionality will be shipping by the end of Q1 with General Availability in Q2.
• CAStor will be sold through ISV’s, application vendors and integrators although we will also work on reference projects with selected end users.
• Caringo has been operating in stealth mode up to now - our website will be available as of 2/15/06
• Please contact mark.goros@caringo.com for information.

From the sound of it, Carpentier and company are on the right track. I look forward to getting the reports back from him (and from at least on beta tester, who I know personally) on how well the product delivers. Interestingly, Carpentier seems to have started some soul searching by other vendors. Diamond Lauffin at Nexsan tells me that his Assureon product is also a software only play, enhanced by some special features built onto Nexsan hardware. Hopefully, they will sell a software-only version of Assureon in the not too distant future.

I wonder sometimes about a phenomenon I have encountered at many client sites: a chief ethics officer. On the heels of highly publicized cases of corporate officer misbehavior, and with regulators in government seeking desperately to find ways to legislate ethics back into the boardroom, many companies have instituted ethics training as part of their standard human resources mission. A few have gone so far as to designate an ethics czar, who, as nearly as I can tell, is the last guy you want to invite to corporate parties and extracurricular events.

On its face, the concept seems ridiculous. Imagine hiring a modern-day Socrates to take your key management out under a tree every lunch hour to debate the nature of the Good? I doubt it would make a company think twice about relocating a polluting plant from a state with tightening environmental policies and restrictions to a state willing to trade its ecosystem for jobs. Clearly, it hasn’t discouraged companies from overlooking the negatives in worker treatment in China in favor of leveraging cheap labor there.

I am inclined to believe that, in most cases, the visibility afforded to ethics is just that — something for show. Rather like my favorite cigarette manufacturer including a URL for a “stop smoking” site in every carton of their product that I buy.

I realize after a recent experience with my eight year old that ethical training begins (and maybe ends) at home. She came into my office while I was making a copy of a DVD onto my hard disk to use with my home theatre system. I have been copying my DVDs to disk to take advantage of a network-based media center that I put in at Christmas and I see no problem with the practice given that I have purchased the physical DVDs that I am copying.

“Is what you are doing legal?” she asked, all wide eyed.

I told her that all I was doing was making a backup copy. I reminded her that her younger brother and sisters had been nabbed a few days earlier taking my movies out of their DVD cases and throwing them on the carpet, where they used them as boogie boards — jumping on them and skating across the room. This put lots of scratches on the media and made a few unplayable. Now, I was making a copy of the disks before they couldn’t be read at all.

She thought about this, then informed me that she always saw international and FBI warnings at the front of her DVDs that stated in no uncertain terms that copying was illegal. I tried to explain that this pertained to folks who were copying with an intention to mass produce and sell copies, while not honoring the rights of the folks who create and distribute the original work. At least, that was my interpretation.

She was satisfied and moved on to other subjects, but it left me wondering. To make my copies of my movies, I must often resort to cloning software and utilities that defeat copy protection or region coding schemes on my disks. It sure feels like theivery, but I know I am in the right.

Then, I drill down into the reality of this issue. Once the technology exists to copy media, the genie is out of the bottle. I have to wonder whether the recording industry will ever be able to stop it.

Sony tried. They shut down a highly profitable disk burner business because it irritated the other side of the house at Sony, which makes video, music and game software. There’s an ethics battle for you: how do you back stringent attempts to prevent piracy, even using root kits, if you are going to sell burners?

They got out of the burner business, making Lite-On an overnight success.

Personally, I think that the movie industry needs to do what the music industry has done with iTunes and the new Napster: leverage the technology. What they have found at iTunes and Napster and a few other pay-per-download services is that folks are willing to pay for good stuff, provided the cost reflects the value they are receiving. Think about it:

Why buy a full CD if only one song is worth listening to? Have you ever felt this way after listening to a new CD by a favorite performer? I know I have. Selective download gives me the option to weed out the crappy tunes and get only the ones that I want. I would pay a few cents per song for that privilege.

Applied to movies, we have another story. My defense of all the folks doing illegal downloads of movies comes down to the quality of the product. I buy certain movies on DVD because I think they are great movies I could watch again and again. More often than not, these are movies I have seen on TV or on long airplane flights, since I don’t go to the cinema very often anymore, so I have already vetted them.

Every once in awhile I will be shown a movie of questionable licensing pedigree (aka bootlegged) by one of my teenagers that they got from a friend of a friend. For the most part, movies today are crap. They should extend the warning scheme from G to X to include special warnings for really dumb plots, terrible acting, low production values, etc. (My lovely wife says that Shannon Tweed, who appears in a raft of movies aired on late night cable, should have her own special warning.) At the end of such a substandard flick, all I can say is, “I’m glad I didn’t shell out even the price of the DVD+R media to record this movie, it sucked so badly.” The good news is that the bootleg let me screen the movie, usually without the same high quality sound and image as a pristine store-bought copy, so I can determine whether I want to buy a copy. Netflix provides another more legitimate method for accomplishing this, but it costs more money.

Bottom line: I find myself a bit at odds with the RIAA and MPAA on the hardline prohibitions they want on disk sharing and copying. Quality, like ethics, should stand for something and maybe the movie makers ought to be willing to tolerate a few copies of their movies being shared in the wild so that the Vox Populi can have a go at them and feed back real information about whether their latest movie is any good. You don’t get this kind of information from sales reports because consumers can’t return a purchased DVD once they have opened it. Thus, sales stats tell you only how many consumers you duped into buying or renting a DVD via hype-based marketing, not whether the movie was any good.

Perhaps the best indicator of quality is how often a movie is requested on an illegal download site, or what kind of seeding ratio it has on bit torrent sites. Movies of any quality — a term broadly used to include idiotic films with hot chicks or guys, and other popular fare — tend to be well seeded. That said, I am confident that good movies will hold their own even in the face of bootlegging. For example, I doubt that the pre-release bit torrent copies of StarWars III made even a tiny dent in LucasFilm’s sales revenues from the completed, digitially-mastered-in-THX rendition of the disk sold legitimately in stores, let alone the massive revenues from theatre tickets, merchandising and everything else that fills the coffers of the 20th Century Fox/LucasFilm organizations. If anything, the bit torrent version helped to build momentum around the movie and to disseminate the view that it was superior to the previous two films in the series and worth seeing and maybe even owning.

So there. I’ve said it. My conscience is clear about copying the disks I own and giving my daughter my simplistic view about backing up my disks when there was so much more to be said.

The previous post on Google’s fight with the US Government not to reveal search info has become a national news subject. CNN and LA Times have done some comprehensive reportage on the issue that is worth a read.

Aside from the privacy implications, which are profound, I loved the LA Times coverage of the huge volume of data that such “search reviews” generate — just one day’s searches fill a lot of storage frames. That’s good for the storage industry, I suppose, but not so good for you and me. Yahoo!, Microsoft and AOL have already knuckled under. Google is the one hold-out. Kudos to Google.

Regarding my response to Aberdeen CEO Bedard, I received a note from him this AM:

Jon,

Don’t be annoyed. I reviewed your website and see that you share a common distaste with the existing analyst options offered to the market. We do 100’s of research studies per year each with more than 150 companies participating. Those are the facts that I reference.

If you think of taking your love for the facts forward and want to work with a great team that shares a common view let me know.

Jamie Bedard

What a guy, that Bedard. Even a flame elicits a marketing pitch. That’s what made America great.

Reported on Slashdot.org today, President Bush wants access to Google search records. Ostensibly, this is to ensure that children aren’t Googling porn, but the reality is probably linked to yet another hairbrained Homeland Security scheme.

The idea, if it succeeds, will dramatically increase the Government’s need for storage. EMC must be wringing its hands with glee. They thought they had a coup when they started shoving Google Desktop on everyone’s PC whose company had deployed Documentum.

A SPAM email signed by Jamie Bedard, CEO of Aberdeen Research, just crossed the transom and received the Toigo Treatment.

Here’s an extract of the email:

From: Bedard, Jamie [mailto:Jamie.Bedard@aberdeen.com]
Sent: Thursday, January 19, 2006 6:17 PM
To: Bedard, Jamie
Subject: Research Editor Career Fair - Media & Publication Aberdeen Members

Sick of writing without the facts to back up your story? Stuck in a job that is just not keeping pace with your career goals? AberdeenGroup is the leading fact-based research firm in the market and ready to help you grow your career!

Over the past two years AberdeenGroup has achieved significant market success with our fact-based research. Did you know that over the past 12 months alone we:

Grew sales and revenues by 70%
Grew our research community to 125,000 members in 40 countries
Completed hundreds of research projects across business and technology
Benchmarked over 17,000 end users on technology and business process performance
Published hundreds of new research reports
Published thousands of new KPIs, metrics, and benchmarks
Launched AberdeenAccess on-line role-based research channels (check it out)
Ready to work in our Boston office and make a difference? Why not, now is the time for a change!

AberdeenGroup is holding an Research Editor Career Fair in our Boston Office from 8:00 A.M. to 5:00 P.M. on Friday, February 3rd. Get a chance to meet the entire team and experience the excitement first hand. To schedule a chance to visit with us please send your confidential resume to Jami.Siler@Aberdeen.com and mention the Research Editor Career Fair in the e-mail.

We currently are holding positions open for all levels of analysts in the following research areas. We will double the size of our team this quarter alone. Why not you?

To which I responded:

Dear Jamie,

Since when have I ever written anything that wasn’t backed up by facts? And why should I believe that your “facts” are facts?

Annoyed,

Jon Toigo

Was I too mean?

For the record, Sid Kilgore is a great guy and a good lawyer (mine). He holds credentials both in the law and in IT and manages to keep a foot in both camps. As a patent lawyer, as an international law specialist, a technologist, and as a friend, I hold him in very high regard.

Last Fall, in fact, I invited him to write a newsletter on compliance for the Data Management Institute to see whether anyone would subscribe to it on an on-going basis. A lot of folks downloaded it, but the subscriptions didn’t come rolling in, which bummed me out.

Today, he dropped another nugget in my inbox. This time it was an article from Law.com discussing how Whistleblower Protection, a mainstay of the enforcement theory and a key social engineering component of Sarbanes-Oxley, is not international in scope. Relationships between corporations with off-shore branches, subsidiaries, suppliers abound in these days of global business. SOX does not, however, reach all of these relationships abroad.

Whistleblowers in, say, the Indonesian or Chinese subsidiary of a US firm may have no protection under the law when they rat out some nefarious financial double-dealing by the SOX compliant mother company in their geo. In practical terms, we have seen the consequences in a recent case:

Ruben Carnero is an Argentinean who worked for two Brazilian subsidiaries of Boston Scientific Corp. as Latin American business development director. In 2002, according to the opinion, he filed suit, claiming that the company had fired him in retaliation for disclosing to the parent company that its subsidiaries created false invoices and inflated sales figures. Carnero v. Boston Scientific Corp., No. 04-1801.

The guy got fired and his termination was upheld by a US Court that argued that protection for whistleblowers did not extend overseas.

So much for SOX and all the pain it has caused in IT producing much of anything worthwhile in terms of corporate governance. Depending on your perspective, this is either proof that you can’t legislate ethics or morality, or that Congress doesn’t do a very good job of riding herd on the guys who fund its members’ campaigns.

Considering the bragging I was hearing from certain storage sales managers for major US vendors when I was in Canada a couple of moths ago, regarding the posting of sales that hadn’t really happened and similar shenanigans, I understand now why everyone was smiling — despite SOX. It is a win win for the storage vendor: fear of SOX has everyone buying storage and storing everything, while lack of fear of SOX has certain vendors doing a lot of underhanded accounting abroad.

Revolutions are needed every once in a while to shake things up. I think the time has come for one in storage management.

I am sick and tired of hearing about good management technology that has been joined at the hip to commodity hardware in order to create yet another proprietary stovepipe solution for a vendor. Are these guys ever going to get it?

• We need an open management solution for heterogeneous environments.
• We need management consolidation, not hardware centralization.
• We need heterogeneous environments because there is no such thing as one-size-fits-all storage. Different applications have different storage requirements.
• We need to tear down the walls erected by vendors around their gear to prohibit cross-platform manageability.
• We need management that is application-facing, not storage facing. Better yet, we need both, if we are to automate resource and service provisioning and data protection.
• And we needed it like yesterday!

Jeremy Evans, head of TPI Technologies, is coming to my house tonight and we will, hopefully, have a site opened by the weekend at www.storagerevolution.com (maybe with a cross link from www.storagedemocracy.com, which is the same thing in my parlance). Modeled on SourceForge and other open source initiatives, we will start building software elements to enable cross platform management and leveraging the workarounds that end users have developed on their own to get the visibility they need into their own storage platforms.

It will be entirely free. Registration will be required to keep the SPAMbots to a crawl, but everyone with an idea to contribute will be invited.

The time has come. The revolution is here.

Don’t get me wrong. I appreciate the efforts of the independent software companies that have been writing management tools for storage. I understand that it has been an uphill battle to get APIs and other connections into vendor gear, if they are forthcoming at all, and I do not hold the ISVs responsible for delivering product to market that can only be as good as the hardware vendors allow them to be.

You have done your best. It is time to do better.

SNIA has not delivered, despite its recent litany of BS in InfoStor this week. SMI-S is an overbuilt joke. It was a decent idea at first and I rather admire some of its early architects. But it depended on vendor cooperation (coopetition) and good will, things that are in increasingly short supply when the market cools and vendors get more cutthroat.

Storage management was never going to be given to us as a gift. That would only happen if storage was a mature industry, which it is not.

In any mature industry, vendors have three fears:

• Fear of standards and their commoditizing impact
• Fear of competitors
• Fear of consumers, who vote with their wallets

The storage industry lacks the third fear: a healthy fear of the consumer. They have not needed, until now, to look over their shoulders to see whether the user is actually benefitting from their stuff or is prepared to jump ship to an alternative…mainly because there have been no alternatives.

If StorageRevolution.com is successful, there will be an alternative shortly — provided that users who are as pissed off as I am with the status quo join in.

Here goes nothing.

Sometimes I have to wonder what folks are smoking…and why they aren’t sharing. A piece I ran across via an email link today, published in Infoworld, is case in point.

Some droid at ESG made some observations that Virtualization (aka LUN aggregation) and iSCSI don’t mix. Apparently, DataCore hasn’t paid the fellow enough to check out their wares.

“If you’re considering iSCSI and virtualization is on your storage road map, however, you may want to think again — at least for the time being. The management tools that take advantage of storage virtualization to aggregate storage subsystems across a network are not yet mature enough to handle iSCSI SANs, says Brian Garrett, lab director at the Enterprise Strategy Group, a market research firm.

The article goes on to suggest that “today’s storage-management software can’t handle iSCSI SANs well. For example, iSCSI doesn’t have a single facility to provide global names, as Fibre Channel does.”

They cite the small install base of iSCSI as a factor in the lack of management and virtualization software.

This whole thing is absolute bullshit. We have iSCSI in our labs. We have virtualized it with DataCore and we have run multiple management packages, backup software packages, data copy packages, etc. across the volumes. I don’t see the problem.

Anybody care to take sides in this?