I just read Steve Duplessie’s take on the HDS acquisition of Bluearc. If I had a dollar for every time I read about HDS buying a NAS appliance, I’d still be poor. They tried it a few times with some small OEMs over the years and in most cases the problem was part technology and part a sales issue.

Let’s face it… HDS is used to selling humongosaur-like systems to the very large enterprises who can afford to buy big iron. Much of HDS’ traditional hardware (manufactured by parent Hitachi) is designed around block-based storage (yes, agree with Steve on this).

Unfortunately for HDS, and lucky for NAS-behemoth NetApp, there are still customers out there, even the big ones, who need file storage  because companies still store a lot of information in the form of files – probably a lot more than you feel comfortable with. I have a 1.5TB redundant NAS appliance at home serving the four members of my family. Yes, applications like ERP, CRM and SCM have limited use for NAS systems and will run a lot faster if the database is running off a powerful SAN engine. But for 100% of employees in any company, they will need to store their files in the network somewhere – and a NAS is a perfect place for that.

So back to HDS… why does HDS need a NAS solution? Likely because customers are hinting they need it. But more importantly lacking a NAS  solution gives competitors like NetApp a window to get in and slowly eat through the HDS armor that surrounds Mr Enterprise customer.

Will this ever work for HDS? I think the bigger challenge for HDS is understanding the technology and being able to sell it convincingly.  From history, this is where the rubber meets the road. This is where all those countless NAS technologies that HDS tried to sell got buried. The good news is HDS has had a few years of history selling BlueArc. Now its just a matter of getting the sales people (in Asia) to get moving.


Some months back I spoke to a senior executive discussing about their product strategy. At one point in the discussion, he quoted an old Arabian Proverb: “the enemy of my enemy is my friend”. It is his interpretation of how partnerships are created in the business world. What scares me is that I understood what he meant and see its practical applications.

A week ago I wrote about SAP’s acquisition of Sybase and its implication to the software arms race. I follow this rant with a twist that comes in the form of new partnerships that are shaping up, not directly as a result of SAP’s acquisition, but as a progression of what’s been happening in the IT industry over the last five years.

And it begins with an interesting article written by Gavin Clarke (the UK Register) around the budding relationship between SAP and HP, as they go after Oracle in a more concerted effort. For HP, its because Oracle dumped its technology (Exadata 1) in favor of Sun (Exadata 2), following Oracle’s acquisition last year of Scott McNealy’s company. And to think that at the 2009 Oracle OpenWorld there was scant anonymity between HP and Oracle despite the Sun relationship.

So with all the knowledge gained from the first Exadata, HP brings a lot of knowledge to bear. Still I think the new SAP appliance will have a tough time competing with the Oracle product (now on Gen 2). SAP will likely integrate its expertise around in-memory architecture (NetWeaver Business Warehouse Accelerator) to market a “powerful calculation engine” for data modeling applications.

SAP Co-CEO  Bill McDermott told Informationweek’s Doug Henschen that SAP had briefed partners including IBM, HP, EMC and Cisco on its plans. Partners are eager to embrace the innovation, he said, and said they will find opportunities to build new products and services around the engine.

Granted that HP, IBM and EMC are interested in building an appliance to compete with the Exadata (with HP having a leg-up at this point in time), at this point IBM would be in the best position given that it has a strong database business, has a captive “mainframe” market, and the resources to spend on R&D. All this without necessarily working with SAP assuming that the German software giant will want to push its Sybase database as part of the bundle. HP may be more interested in this although it would hurt its relationship with Oracle big time (we’ve seen this with HP’s relationship with Cisco).

Speaking of which, a Cisco-EMC-SAP triumvirate may present the best opportunity to run against Oracle.

For the moment, Oracle has the upper hand, having created an opportunity with the Exadata and the rest of the industry is playing catch-up. Larry Ellison is having a ball.

Streetinsider.com recently posted its commentary following IDC’s 2010 prediction that IBM would acquire Juniper Network. This may be a response to HP’s ongoing acquisition of 3Com and the continuing speculation of a potential Cisco-EMC merger.

The much delayed completion of the acquisiton of Sun Microsystems by Oracle will likely come to conclusion during the first of 2010.

Seeing itself more a bridesmaid than a bride is Brocade, which has put up its “For Sale” sign and seen no real suitor moving to take the offer. As far back as October 2009, Gigaom analysts pointed to Juniper as a potential buyer for Brocade. IBM would have also been a good suitor but the overlap is in the Foundry products (assuming Big Blue does acquire Juniper).

I still think Brocade would made a fine addition to HP. Of course Brocade may need to get rid of its Foundry portfolio to minimize overlap with HP’s Procurve offerring.

Assuming that Cisco and EMC do combine, that would leave NetApp and HDS as the lone pure play storage vendors. Can these two survive in an environment where customers want true ease of use, best possible integration, and one throat-to-choke accountability from their vendors?

If you look back at all this a proliferation of vendors came about because customers clamored for more choice. But while enterprises say they like choice because it allows them to haggle on the price, in reality, they will turn to one vendor who has everything because its simply easier to deal with one supplier than a multitude of vendors.

Having a choice of sources is great only in as much as it gives you the perception of freedom to pick what you want. But choice brings chaos to the equation. This goes agaiinst the grain of businesses that must operate in an orderly fashion so that processes can be streamlined resulting in greater efficiency, higher productivity and lower total cost of doing business. Which if my math is correct will mean greater profits and better shareholder value.

So the old adage of vendor lock-in doesn’t hold sway in today’s highly competitive world. As one CEO told me recently, “At the end of the day, its really not a matter of giving customers a wide array of choices to choose from. It is giving each customer the product that he or she will want to buy. And that could be just one product it just so happens to look a little different to the individual customer.”

Smoke and mirrors!

Picked up early this morning on Wall Street Journal is breaking news about a potential sale of Brocade to the right buyer. Among the parties interested in the networking vendor are Hewlett-Packard and Oracle.

Brocade has a market cap of US$3.2 billion with 2008 revenue at nearly US$1.5 billion.

Any sale of Brocade will likely impact the business of HP, IBM, HDS, Sun Microsystems and EMC. All these vendors resell Brocade-made products either under OEM or the Brocade label. In the short-term, Cisco sales people will have a field day running after long-time Brocade customers and channel partners as the usual “fear, uncertainty and doubt” or FUD gets thrown in to wreck the nerves of companies that have invested heavily in Brocade technology to keep their storage area networks up and running.

Brocade dominates the Fibre Channel switch market following its acquisiton of McData in 2006. Recognizing the importance of Ethernet in the overall network storage fabric ecosystem, it bought Foundry Networks in 2008.

Rumors of Brocade being up for sale dates back as far as 2003 when McData was rumored to be in talks with Brocade to acquire the latter. But three years later, it was the other way around. I doubt that we will see Cisco buying Brocade if only to kill the competition and dominate the market. but certainly if HP acquires Brocade it would significantly enhance its networking capability.

Rumors of HP’s interest in Brocade have been floating around for years. It would certainly complement’s HP’s networking portfolio which is largely hinged on the HP ProCurve business and supplemented by its OEM deal with Brocade.

Oracle is the bigger wildcard here. Oracle is in the midst of closing its Sun Microsystems acquisitions. The market is awash with rumors that the hardware piece of Sun would be sold off. But if Oracle were to acquire another hardware vendor – say Brocade – it would certainly mean that any potential sale of Sun hardware may no longer be on the table and that Oracle is really commited to its CEO’s vision of offering customers complete systems.

Ittai Kidron, an analyst at Oppenheimer & Co., explains in a research note that an Oracle acquisition of Brocade suggests the company will have to commit to the hardware business for the long-term. Brocade is a nice fit and has no overlap with its Sun purchase according to Kidron.

Any acquisition of Brocade by any of the server vendor will have to be thought out properly. Brocade has large OEM and reseller deals with a number of server vendors. As the Cisco entry into the server business has shown, vendors will more than likely shop elsewhere if Brocade becomes just another product line of a server vendor.

Enterprise Strategy Group analyst Bob Laliberte reckons IBM and Dell may also be potential buyers, “I don’t think you’ll see EMC or Cisco buy them.”

Stifel Nicolaus Equity Research analyst Aaron Rakers also put in Juniper as another possible buyer.

Watch this space as we monitor, report and analyze the developments surrounding the potential sale of Brocade. At this point it is everyone’s guess as neither Brocade nor the potential suitors are saying mum. It is also very possible that Brocade is just trying to get a feel for the market. Afterall rumors are the stuff of the tech industry.

Encryption is the age old approach to making data unreadable to anyone but the intended recipient. While originally created for military purpose, it has found commercial success not long after the war when companies and individuals needed to protect development work from competition – at least until it gets patented.

Encryption is used to protect data, whether it is in-transit or at rest. In 2007, the US government reported that up to 71% of companies surveyed use encryption for some of their data in transit.

Approaches to encryption vary by vendor. Some encrypt data as it travels between sender and receiver. Encryption for data at rest is also common, particularly as portable computing devices have become a normal business tool.

Hard drive vendors have been working to introduce encryption into drive itself. My first introduction to this was while attending a Fujitsu Expo in Tokyo. One of the displays was a Fujistu 2.5” hard drive (HDD) that uses a built-in encryption technology to protect data stored on the drive. .

Recently Seagate and Hitachi launched their implementations of encryption on their HDDs. Seagate uses what it calls Secure Self-Encrypting Drive (SED) option across its enterprise-class HDDs. Hitachi Global Storage Technology (GST) calls their technology “Bulk Data Encryption” or BDE.

Hitachi’s BDE technology is based around the Advanced Encryption Standard (AES 128) supported by the National Institute of Standards and Technology (NIST). Seagate uses that security protocol developed by the Trusted Computing Group, a non-profit industry body group focused on developing open standards.

The advantage of having data on the drive encrypted is evident whether it involves desktop computers or laptops. For desktops, if the company decides to upgrade their PCs, destroying the electronic content of the HDDs is important. Because of the potential to get stolen or lost, laptops need the disk encryption.

HDDs used in enterprise-class arrays have one characteristic unique to their class – they are purpose-built to continue operating 24×7. With the exception of COPAN’s implementation of MAID (Massive Array of Idle Disks), the HDDs’ on most arrays continue to spin throughout their lifecycle. They are only taken down when the array firmware indicates a soft or hard failure. In this instance, engineers often take out the faulty drive and sends over to the manufacturer for testing. But failed HDDs still contain the data.

According to Michael Willett, co-chair for the Trusted Computing Group Storage Work Group and Seagate Research, over 50,000 hard drives, thought to be extremely safe within the data center, are decommissioned and leave the data center daily. For returned drives with suspected problems, an IBM study indicates that 90 percent are still readable, allowing unauthorized personnel easy access to confidential data. Classifying the data on any decommissioned drive as secure without taking the proper security steps could lead to a data breach situation.

In the past vendors like EMC and Dell have clauses in their contracts that allow customers to keep faulty drives on their premises. This may no longer be necessary.

By having the encryption built-into the HDD hardware, data written on the disc platter remains encrypted and unreadable even if the HDD is repaired. Now if Mr. Chen had this on this laptop when he sent it for repair, the technician wouldn’t have been able to pry open the data even with a crowbar. That ladies and gentlemen is how you keep curious eyes from knowing your secrets.

For a listing of encryption software, click here.
Disclaimer: I don’t claim to know the software on the list. I suggest you walk with caution.

Ok, so I swiped that off from the original “water, water everywhere but not a drop to drink” Samuel Taylor Coleridge poem “The Rhyme of the Ancient Mariner” in which sailors on board a ship where without water to drink despite the fact they were surrounded by a sea of water.

In the same token, companies have successfully figured out how to capture as much data as they could ever need. The problem is interpreting the data to mean something.

One industry that is trying to cope with this issue is the “legal” industry where until recently the fight in the ediscovery solution suite was pretty much among small niche layers.

As many should know by now storage gargantuan EMC took a bite of the ediscovery market with the acquisition of Kazeon. EMC said the acquisition gives it an “end-to-end, in-house d-discovery and litigation readiness solution”.

This is bad news for Kazeon competitors, Iron Mountain, Autonomy, Clearwell Systems and StoredIQ. The latter because it is an OEM partner to EMC.

The two competitor that just might have to start shopping for an OEM partner in this space is NetApp, which partnered with Kazeon in October 2005.

Is there a pattern here?

On May 20 2009 NetApp offered to buy Data Domain (DD) for $25 per DD share. It all seemed a natural fit. NetApp needed a robust data deduplication solution that fits well with its target of SME (ok, more the mid-market than the small enterprises – but definitions of what constitutes a SME varies by country or region). NetApp acquired Alacritus in April 2005 for its virtual tape library technology. Speculation abounds this will be thrown out in favor of DD’s technology.

For its part, Data Domain needed the global reach that NetApp had to offer.

Greg Cornfield noted that this would give EMC (with its acquisition of Avamar in 2007 for $165M) a run for its money (http://www.searchstorageasia.com/content/netapp-buy-data-domain). Other industry bloggers

On June 1, EMC did up the ante with EMC CEO, Joe Tucci, holding informal talks with Frank Slootman, Data Domain CEO, followed by a letter of offer. The $30 per share cash offer by EMC certainly is more appealing than the cash + stock offer by NetApp.

But the story doesn’t end there. As with any romance novel, there is a twist in the plot and a counter offer by NetApp followed the EMC bid.

Why the interest in Data Domain? DD’s main claim to fame is its leadership positioning in the data deduplication business. NetApp desperately needs it to enhance its offering. EMC could do well to add the strength of the DD technology to its portfolio of offerings. EMC’s current offering is a combo of Quantum (OEMed) technology and the Avamar acquisition.

Gartner estimates that about 30% of companies have deployed some form of data de-duplication. Data deduplication allows companies to reduce data storage requirements anywhere up to 30:1. The dollar savings appeals to almost everyone with a growing storage farm regardless of prevailing market conditions.

Gigaom’s Stacey Higginbotham thinks EMC will fight back. EMC is doing this to protect its turf and stop a rival from eating into it’s (EMC) market.

Ryan Hutchinson of Lazard Capital believes that EMC is doing this as defensive as well as offensive. EMC has the cash to buy Data Domain at a price that would certainly hurt NetApp (even if the latter is the final winner). EMC has a cash hoard of $9B (cash plus long-term investments) relatived to NetApp’s $2.6 billion (cash and short-term investments).

At any rate, the discussion appears to be mute at this point. Data Domain and NetApp have published a new press release to indicate that they have “entered into a revised acquisition agreement” valued at $1.9B or $30 cash and stock.

Is the deal over? Outside of approvals and due diligence – maybe. Would EMC counter back? They certainly got the money. Broadpoint AmTech analyst Brian Marshall seems to think EMC can still come back to the table with a higher bid if only to keep competition from getting their hands on DD’s technology.

What’s weird about all this is that NetApp’s share price is about US$18 compared to around US$32 for Data Domain. For the life of me I can never figure out how stock valuations work.