Terms like “software-defined storage” (SDS) can be misinterpreted as representing something that came and went…some may even view it as a case of ‘failure to launch’; but the truth is that it is more like a rocket that had a very long and slow-burning fuse, but is now gradually escaping the gravitational pull of old-style IT.
Last week I was pleased to be asked to host a couple of panels pulled together by “Datera and Friends” and held at HPE Discover. The ‘friends’ across the two panels included Datera and HPE (obviously), Intel, Mellanox, SereneIT (a channel partner), Morae Global (a customer), Cumulus, Scality and WekaIO. You can watch the panels Leading the Digital Transformation to SDDC and Software-Defined Everything online now – I commend them to you, but just make sure to grab a coffee as each runs a little over an hour. The value of these is that they are not the normal 15 minute multi-faceted thinly-veiled promotional platforms: Instead they are long-form discussions to try to really get to grips with the topic…what is SDS (or SDx, or SDDC, or Digital Transformation, or Composability, or the myriad semantic treatments of a more flexible approach to IT), and what is its value?
HPE Discover was a fascinating backdrop for the discussion, because HPE (like many a large systems vendor) finds itself having to straddle the IT fence: a slew of new product announcements -spanning cloud, storage, data management, HCI, analytics, networking and services- on the one side; and a commitment to impending ‘as a Service’ options, yet balanced with a ‘Cloudless’ vision, on the other. As mentioned in ESG’s On Location video of the event, despite the balancing act this seemed to be a year of HPE turning the corner. But like any big vessel it will take a long time to complete the maneuver. Some of the elongated time is because having a foot in both the traditional and the contemporary camps is the bane of vendors like HPE (it’s also how they make their billions!), but it does demonstrate a degree of consensus around what contemporary/future IT will look like.
And it was that consensus around the optimum contemporary journey to our best vision of the data/IT future that these panels were focused upon. The range of panelists represented something of a first effort at an implicit (I, for one, hope it becomes real) coalition around the topic and approach. Certainly, the old adage that “a rising tide lifts all boats” is applicable whenever erstwhile competitors work together; and when – as here – you are discussing and embracing fundamental change that threatens some mighty entrenched forces (which I mean in both senses!), co-operating to reach a critical mass and volume is an inherently good idea. After all, history is full of better ideas that got beaten out by marketing and distribution muscle…
In the video panels you will see a number of themes coalescing: the decoupling of hardware from data and apps (with automation as a prerequisite), open-ness, “Service-defined”, dynamic agility, doing things both faster and better/differently, and all adding up to system-(aka business-need-) defined infrastructures. This is not just word soup – it is more about encapsulating a different approach to IT. It is, if you will, about getting back to IT as Information Technology and away from it having been constrained to be Infrastructure Technology for recent decades. It is not independently about the particular offerings of Datera or anyone else on the panel, and that’s why the discussions were valuable. After all, pretty much everything in IT (irrespective of which definition of that term that you prefer!) is fungible. Data is not. And so what all the panelists were discussing – what they have coalesced around, even if not an eventual all-encompassing name! – is data fluidity.
When expressed like that, what was being discussed was far from chatting about the future of storage (however elegantly that is done, it remains a fairly static necessity), but was instead an erudite investigation of how to optimally enable the future of IT. That future must, of necessity, be founded on data-service-freedom. It is, if you like, akin to “Data as a service” – whether ‘destinationally’ that’s on premises, in a cloud, or whatever comes next, software-defined storage (to use our semantic starting point) is not a destination but an enabler for a better way of doing IT.