By Chris Mellor, Contributor
The Vblock phenomenon of integrated stacks has begun to make its presence felt, and we're seeing a wave of storage vendor consolidation, with Dell, EMC and IBM all making significant recent acquisitions. These vendors want to consolidate to improve their offerings in a world in which customers are seeking storage efficiency by virtualising servers, buying storage that supports both file and block access storage, and starting to store data in the cloud.
Customers are also beginning to look at buying converged IT stacks because of the cost and management efficiencies they offer, with the VCE coalition and Oracle leading the way, closely followed by NetApp's FlexPods and HP. Dell is also preparing converged IT stack offers, as is IBM, which bought Blade Network Technologies for its Ethernet technology.
But, there are still some holes in these companies' converged IT stack offerings and yet more holes in the storage offerings of a second group of companies, including Hitachi Data Systems along with the aforementioned NetApp and Oracle.
There has also been consolidation in the storage networking supplier group, where Ethernet/FCoE technology has been the main focus for Brocade, Emulex and QLogic. Meanwhile, LSI bought StoreAge for SAN virtualisation software and OnStor for filer technology.
The tape market has also seen consolidation with Quantum buying ADIC and Tandberg Data buying Exabyte and its VXA tape technology, though those acquisitions date back to 2006. More recently, Overland Storage looked for expansion beyond the tape market by buying the Adaptec SnapServer NAS business plus Quantum and MaxiScale clustering assets.
It's true there have been losers in this storage vendor consolidation era. No one wanted mainstream business optical storage and Plasmon crashed, while holographic storage startup efforts came to naught. Some filer companies also crashed, such as Exanet and OnStor. File virtualisation dribbled away with F5 buying Acopia and EMC's Rainfinity purchase failing to set the world on fire.
In-fabric SAN storage virtualisation has stabilised with IBM's SVC dominating it and HP OEMing the StoreAge technology acquired by LSI. No one else is looking to pump product into that space now.
What does this mean for the standalone storage supplier survivors? Are they now also-rans, effectively banished from the top tier of storage suppliers? I'm thinking here of BlueArc, DataDirect Networks, ExaGrid, Nexsan, Panasas, Pillar Data, Sepaton and Xiotech, among others.
For most of them, their future is their present; where they are now is where they will stay. Startups will go public if they grow fast enough, but it is unlikely, in my view right now, that any of these companies will be bought by established suppliers -- with three exceptions: BlueArc, CommVault and Xiotech. All three have distinctive technologies that give their products an edge and/or a unique identity. BlueArc is heading toward an IPO and, once that is over, may well reveal new strategic directions as it accepts that its goal is not to be absorbed by a larger supplier such as Hitachi.
How will BlueArc respond to the cloud? Could it provide a very high-performance cloud gateway using its own technology, some flash pizzazz and Aspera network acceleration?
CommVault is steadily growing and extending its capabilities as a storage managing, protecting and archiving provider. It has strong relationships with companies like NetApp and Dell and could carry on growing as an independent or be snapped up. Such an event would give yet more grief to Symantec as another one of its major partners gains the technology to compete more strongly with it.
Xiotech seems to be crying out for integration with upper-storage-stack software or a hypervisor storage management layer. Were Symantec to get directly into the storage hardware business, then Xiotech would be a good fit for its Storage Foundation software. It's possible that Xiotech is so unique in its technology and approach that it will forge its own way, go to IPO and become a stable and successful standalone vendor.
Pillar Data has terrific technology, but there is a huge unspoken elephant in its room in the shape of Larry Ellison. Oracle's purchase of Sun seems to have closed off the possibility of Oracle absorbing Pillar. So its future, like Xiotech's, looks to be an IPO based on earning big dollars from building better storage mousetraps, adopting flash and cloud gateway technology in the kind of innovative way we've come to expect from Pillar.
Niche data deduplication suppliers like Sepaton and ExaGrid and substantial storage software management and protection suppliers like CA and Symantec will continue. They don't appear to be acquisition targets and don't seem to be acquirers either, at least not in any large way. The same goes for NEC and Fujitsu, although the latter company is still evolving as it digests Fujitsu Siemens Computers.
There has not been storage vendor consolidation in the main new technology activity areas: flash memory and cloud storage, which have both seen substantial startup action. In flash we have Fusion-io; SanDisk, which resisted a Samsung acquisition attempt; STEC; Violin Memory; Texas Memory Systems; and newer startups like Anobit, Pliant and SandForce.
Filer acceleration company Avere is also a flash technology user as are cloud storage gateway startups like BridgeStor and StorSimple. Other startup companies active in this space include Aspera, Cirtas and Nasuni, as well as established WAN acceleration player Riverbed with its Whitewater product.
The two new storage technology frontiers are flash and the cloud, which build on storage virtualisation and high-performance unstructured data access -- both ongoing areas of product build-out by suppliers. The two new storage business frontiers are converged IT stacks, as mentioned above, and the cloud.
Suppliers intent on competing in the converged IT stack space need to create sharply defined bundles or templates modelled on the VCE coalition's Vblock -- as are NetApp/Cisco's FlexPods -- or they need to bring together in-house components. Dell looks to be doing this, although it still needs in-house networking, and so too is IBM with a blade server/V7000/Blade Network Technologies combination. Hitachi Data Systems has access to Hitachi servers and networking alongside its storage products, and HP already has most of the hardware ingredients it needs.
Jumping into the prediction game I'd say Dell could buy networking technology to fill its biggest in-house converged IT stack hardware gap. Beyond that we'll maybe see in a year or two a rush into the cloud storage gateway area by our four main consolidators, Dell, EMC, HP and IBM, partly prompted by a probable entry into this area by one or more cloud storage service suppliers such as Amazon, Google, Iron Mountain, Microsoft Azure, Nirvanix or Seagate's i365. What better way to ensure business customers use your cloud than by offering them a gateway to it?
About the author
Chris Mellor is storage editor with The Register.
This was first published in December 2010