These three haven't experienced massive growth. Nor are they in the right market niche to have been scooped up when major suppliers faced a product gap, as was the case in the iSCSI SAN space with EqualLogic going to Dell and LeftHand Networks to HP.
Nexsan, Pillar Data and Xiotech don’t look to be in a position where they could receive more funding, or have an IPO (with the possible exception of Nexsan) or be bought. Instead, the foreseeable future appears to be one of organic growth with backers cashing out what they can by way of profit when growth has been strong enough to warrant an IPO.
Are there common factors that unite these three companies, or are they three special cases that for one reason or another are strong enough not to have failed, but not lucky or brilliant enough to have been bought or gone for IPO yet? Let’s explore the issue.
Nexsan competes with the Fibre Channel SAN storage array vendors. The differentiation Nexsan offers is the claimed cost-effectiveness and density of it SATAbeast and SATAboy arrays. Nexsan grew well enough, despite the competition adopting some of its technology, to encourage its management to try for an IPO in 2008.
But market conditions as recession loomed were inauspicious, and Nexsan's history and product range were not strong or differentiated enough, and the IPO was pulled.
Two years later Nexsan tried again for an IPO. By then it offered spin-down in its arrays and positioned them as energy-efficient disk archives, like those from Copan (which subsequently crashed and burned). Once again, market conditions were against it, and the IPO was cancelled in April 2010.
Now Nexsan has rebranded, with its E Series arrays aimed at midmarket customers, and it is growing again. One train of thought is that Nexsan has tried to boost margins by positioning itself as a premium low-end or midrange supplier but that this has opened it up to competition from the likes of Infortrend and Promise, thereby limiting Nexsan’s market reach. We shall have to see how this plays out as it tries for an IPO again this year.
Pillar Data, the creation of IBM storage veteran Mike Workman, developed an innovative three-component architecture, with storage Bricks, I/O controller Slammers and managing Pilots in its Axiom SAN arrays. The radical idea here was to short-stroke the drives, positioning data that needed fast access on the outer tracks of the drives and minimising read/write head seek time while filling the inner tracks with less active data.
Thus Pillar could offer different qualities of storage service from the same set of disk drives. The Axiom software was developed to extend this, adding in data protection and high availability, for example, and new hardware storage tiers were added to create a powerful offering.
Pillar competed with 3PAR and Compellent, and the rough positioning was that 3PAR’s InServ overlapped the Axiom but went higher up the enterprise storage ladder. Compellent also overlapped Pillar but went lower down the SMB ladder. These two Pillar competitors turned out to be well-positioned, with Compellent’s products matching gaps in Dell's range when it wanted to break free from EMC, and 3PAR’s providing a substitute for a tired HP EVA line and the OEMed XP from Hitachi.
Pillar has an enthusiastic and loyal customer base, as well as a good development programme for its Axiom hardware and software. But, life has become harder as the HP and Dell channels now push competing products, though Pillar is not daunted and should continue growing. And don’t forget, Pillar is financially backed by Oracle’s Larry Ellison with the possible outcome that it is folded into that empire to fill a need for second generation SAN storage in the Sun stable.
Xiotech is the king of sealed canister disk drives. Its radical idea was to group a bunch of disk drives together and manage them with sophisticated Seagate-derived drive software that could recover apparently broken drives and keep them working. It provides a closed enclosure, called the Intelligent Storage Element (ISE), a single logical disk that has a five-year warranty against failure. This means a storage array built of ISE storage blades is inherently more reliable than traditional SAN arrays and doesn’t need a conventional storage array controller.
ISE is a good fit for higher-level storage stack software products such as Storage Foundation from Symantec and DataCore’s SANsymphony. SSD support is being added to improve performance, which was already good, and ISEs are appearing in blue-chip enterprise data centres. Now there is talk of a partnership with Sanbolic with its Melio FS SAN file system layered atop Xiotech's ISE storage blades.
Xiotech’s ISE technology came when it acquired the technology and product group from Seagate. We can speculate that Xiotech, despite having good technology and product, has not been the target of acquisition attempts by mainstream storage array vendors because its technology is too disruptive. Xiotech has to grow its business itself, without the accelerative influence of an acquisition.
This is what links it to Nexsan and Pillar Data. The three were not snapped up in the recent few years of storage acquisition enthusiasm because they did not match up with any potential acquiring companies’ product and market gaps or weaknesses, not because their technology was inferior.
In a storage market bestrewn with casualties and failures, they have survived and grown and are now making their own way.
Their main differentiation is that these three storage array vendors provide best-of-breed product in the storage market niches they sell into. That's their challenge, and as long as they keep growing their business and developing their products to keep up with the trends affecting their customers, they should be set fair.
This was first published in April 2011