What’s in a name? Would storage by any other name perform as sweetly? Are all buzzworded products equal? In tech, buzzwords are rendered meaningless very quickly, and empires are more often built on being first to market, rarely on being the best on the market.
Hyperconvergence is a fantastic example of a tech market whose primary defining buzzword has been stretched, tortured, mangled and abused to the point of meaninglessness. Originally, hyperconvergence meant lashing together all the hard drives and SSDs inside a virtualization cluster into centralized storage. This allowed critical functions such as VM vMotion/Live Migration to occur without needing an expensive SAN or enterprise NAS.
Eventually, things got a little bit fuzzier. Companies started to claim hyperconvergence if they had cluster-wide server side caching or storage gateways that presented to the hypervisor like they were a hyperconverged storage offering, but still utilized one (or more) centralized storage units to provide most of the storage.
As the definition of hyperconvergence was stretched to accommodate every marketing department that wanted in on the buzzword, it seems everyone (with a couple of notable exceptions) has brought to market a hyperconverged solution.
Nutanix was founded in 2009 and shortly thereafter launched it’s first hyperconverged node. As the lore goes, the term “hyperconvergence” was coined at the VMworld where Nutanix came out of stealth mode in 2011. (It should be noted that Pivot 3 and a few others were offering practical hyperconverged products for years before the term was coined.)
Unfortunately, in the nearly five years since hyperconvergence became “a thing,” it seems that very few of the vendors occupying this space have cared to look beyond “storage + compute = novelty” and deliver something that actually advances the state of the art or solves real world customer problems.
That’s about to change.
Feature versus product
The storage industry is fascinated by the concept of storage for storage’s sake, but storage of any kind is a feature, not a product. Businesses don’t care about the storage that powers their datacenters any more than they care about the brand name of the oil used in their fleet of cars. If it does the job adequately and for a price the organization is willing to pay, then it should be out of sight and out of mind.
By this logic, if storage vendors want to stay relevant and be anything other than a commodity, they should be focused on incorporating other features by default. When we buy virtualization clusters we should be seeing self-service portals, API-based infrastructure-as-code provisioning and management, role-based administration, templates and recipes, and more besides.
In short, today’s storage vendors should be leaning towards the infrastructure endgame machine model, rather than wasting everyone’s time trying to convince us all that hyperconvergence is a product instead of just a hardware configuration; and a fairly limited configuration at that. This model renders the hardware infrastructure invisible to the user, and manifests itself as a software defined infrastructure platform where storage + compute + network + management & all other goodies = infrastructure endgame machine.
Yottabyte is well positioned to deliver exactly this. We have been running our own cloud for years. Our software has evolved to meet the real world needs of our customers and by now is more than proven. YottaBlox, powered by the yCenter SDI Platform, offers everything needed for a “cloud in a can” solution.
Yottabyte is ready to take the industry beyond hyperconverged hardware. We’re ready to stop pretending that storage is a product and deliver the features required to meet business needs.
Join us @YottabyteLLC, @greg_m_campbell & @dduanetursi at Intel’s Cloud Day 2016 (@IntelITCenter | #IntelCloudDay | #NowPossible) and let’s take that next step together.