In a dramatic move this morning, Dell announced the intention to purchase the innovative vendor of enterprise storage 3PAR for $1.15 billion in cash. If there were any doubts remaining about Dell’s commitment to be a force in the storage market alongside of EMC, IBM, HDS et al., this deal should put them to rest. Dell acquired the iSCSI storage array vendor EqualLogic in November of 2007, clustered NAS vendor Exanet in February of this year, and most recently they bought data deduplication vendor Ocarina this past July, as well as putting together a partnership with object storage vendor Caringo. Clearly these are a significant list of deals, but the strategy was incomplete without an enterprise class primary storage system of their own. 3PAR, whose products generally compete with high-end systems in terms of performance and availability, will give Dell the ammunition they need to go head-to-head with the big guys.
Dell has cultivated a relatively successful partnership with EMC for mid-range and enterprise storage for some years, but in spite of Dell’s claim to be invested in that relationship going forward, this deal clearly puts pressure on it. Initially, there is a gap between the SMB focused EqualLogic products and the high-end offerings from 3PAR, which will be filled by the Clariion products from EMC, but in the long run, Dell is likely to be motivated to move out of EMC’s shadow and build its storage brand on proprietary products based on these acquisitions.
This deal ends a good deal of speculation about who might buy 3PAR, with HP the main alternative suitor. HP now faces a build or buy decision as it continues to try to redefine itself in storage amidst a patchwork of the aging EVA platform, partner technology from Hitachi on the high-end, and acquisitions in iSCSI and clustered file storage, but no clearly defined long term vision or anchor technology.
I've been down on vendors pitching half baked cloud storage visions for some time now, which often seem like attempts to avoid being left out of a feeding frenzy, rather than a vision of how the cloud might improve real storage environments. So I've been fairly negative on the hype of the cloud, and the data seems to support this negativity. According to Forrester's 2009 Enterprise And SMB Hardware Survey, 91% of respondents say they have no tangible plans to move forward with cloud storage. But when I stop to consider it, I don't want to look back in 5 years and say that I was the analyst that said cloud was never going to happen. There's a huge amount of pain in enterprise IT, and storage is often the most problematic segment in the data center. It has ever-increasingly high costs, challenging hiring of the right skill sets, and poor efficiency. Chaos is becoming the norm. The prospect of getting effective and secure services from an outside band of experts or of building out a consistent, measurable, and scalable internal architecture resonates strongly with IT buyers. So, there's big pain, budget funds, and a vague vision of an alternative solution. That's a recipe for eventual change if I've ever seen one.
On July 19th, Dell announced their intention to acquire data deduplication vendor Ocarina Networks. It’s no surprise that somebody bought Ocarina, as deduplication is one of the hottest technology areas in storage today, with every vendor scrambling to offer tools that can help users contain their massive growth of data and storage footprints. Data deduplication is a form of virtualization that breaks data into chunks, and then compares them and eliminates repeated chunks, replacing them with a pointer to the originally seen version.
NetApp has led the charge in primary storage with their complimentary NetApp deduplication feature within their Data ONTAP operating system. EMC made a big move into dedupe with their acquisition of Data Domain, mostly used for data backup. So no wonder that competitors of NetApp and EMC are eager to get something going in this hot space. And no wonder that Ocarina was an obvious target with their advanced deduplication algorithms that can eliminate redundant data even in image files with similar coloring, often seen as a particularly tricky file type for deduping. Also, Ocarina is based in an appliance, which means that it can be applied across a whole environment, rather than only on data within a single storage system. This type of functionality is likely to be core to enterprise storage efficiency strategies going forward, and is probably better served up by a large vendor that sells their own storage, rather than as an add on from an unknown quantity.
Now that the dust has settled on the Oracle acquisition of Sun Microsystems, I have to say a bittersweet farewell to the first global tech firm I crossed paths with as a young intern. I have to say that as an analyst over the past three and a half years, I had a tough time working with Sun Storage: the patchwork portfolio, the revolving door leadership style, the $4Billion dollar oil and water acquisition of StorageTek and deeply questionable corporate marketing efforts.
Today, HP announced that they will be acquiring the startup cluster file storage software firm IBRIX in a deal with few disclosed details. Congrats to HP for making another interesting acquisition in the clustered storage space, following their Feb. 2007 acquisition of Polyserve and their October 2008 acquisition of Lefthand. IBRIX makes software that allows users to cluster many nodes of storage (either industry standard servers or iSCSI/FC SAN arrays) for use in high performance global namespace file storage.IBRIX’s customer base includes some 175 firms, generally in the high performance compute (HPC), video rendering/animation, and oil and gas exploration spaces.
I care deeply about the environment, certainly more than I care personally about money, so it pains me to say that in most cases, making storage decisions based on power expenditure alone is not rational behavior. The world is driven by economics, and the stark reality is that the cost of power is only a drop in the bucket compared to the amount organizations spend on acquiring and managing their enterprise storage systems. Maybe someday a consumption tax or cap and trade system will tip the balance towards more responsible consumption of non-renewable resources, but in the meantime, the pricing of power (especially in the US) doesn’t give much economic incentive for good behavior. In fact, according to a report Forrester published recently, the amount of money typically spent on electricity to power and cool a TB of storage is only about 1% of the cost of buying that TB of storage (or about 4% of the annualized cost of buying that storage given that you only have to buy the TB once every 3-5 years but you power it every year).
Today, HP announced the intention to acquire iSCSI storage vendor LeftHand Networks for $360 Million in cash. LeftHand makes virtualized storage array software that can turn just about any server, PC or subset of disks on a device into a full featured IP based SAN array with features like space efficient snapshots, distance replication support and thin provisioning. LeftHand has a tight partnership with VMware, and does both standard virtual server attach to SAN, as well as some innovative virtual appliance models that turn the physical disks on board a VMware server into an array. The LeftHand software can span across multiple physical devices, allowing for combinations of virtual appliances and dedicated servers.
Yesterday at Oracle OpenWorld, Larry Ellison announced the database giant's first foray into the hardware realm, unveiling the HP/Oracle Database Machine, branded as Exadata. The Exadata system is a combination of Oracle's 11g database engine using Automated Storage Management (ASM) to manage a grid of HP Proliant servers with 12 SATA or SAS drives each, connected to database servers via InfiniBand. Oracle separated the database processing to allow parts of queries to be run on the storage servers, delivering query results to the database servers rather than whole tables, reducing the amount of I/O passing between servers and storage. There is no interconnect among the storage servers, instead ASM manages read and write requests in parallel among all the storage boxes. The storage server grid allows replacement of failed nodes and ongoing scalability to be automated, and growth to be theoretically unlimited. Oracle claims significant performance increases over traditional architectures and dramatic price improvements at the same time, given the industry standard x86 server architecture and the wide use of dense SATA drives.