A Vision of the Future Data Centre – Digital leadership

Picking up where we left off on our hyperconverged journey we can look to history which has revealed to us a natural evolution of innovation.  It starts with invention, progresses to custom builds, then to product and finally commodity (see Simon Wardley’s @swardley articulate descriptions in his blog posts).  The evolution of the data centre is no different.

In the late 70’s the stranglehold of the mainframe vendors was broken with the advent of the open system, and for the next three decades multiple vendors produced competing products which worked together to deliver IT solutions. But steadily, limitations within the design of the traditional architecture came to bear, so each layer became commodity – differentiated mainly on price.  Software defined data centres and hyperconverged infrastructures represent the start of a new line of evolution – one that offers more value than its predecessor. So, where might it take us?

The hyperconverged solutions which are gaining so much attention provide an excellent solution which alleviates many of the issues that have plagued IT departments for the last 35 years. The software defined solutions created by Amazon and Google show that an architecture of x86 servers designed can scale way beyond the needs of nearly all businesses, and through organisations like Nutanix and Simplivity  these solutions are now obtainable by organisations without R&D budgets that compare with the GDP of a small country.  But doesn’t all that evolution threaten to bring us right back to where we were 35 years ago – locked-in to a single vendor providing compute, storage and network from a single box?  I think not, because of the relentless drive towards commoditisation

Hyperconverged solutions represent a major step towards the vision described above, yet the vision I have is much more commoditised.  The future I see is one that is truly software defined where any x86 architecture can be added to the stack.  To avoid becoming the exclusive domain of a single vendor, the software bit in ‘software-defined’ will need to be open-source, which while VMware’s solutions are extremely slick, they can’t fit that vision in their current format. Commoditising the core of the IT infrastructure allows effort and expense to be spent on what really matters – innovation at the edge. The opportunities opened up through proliferation of mobile devices, wearable technology, cloud and big data is where change is really happening – change that impacts our lives both professionally and personally.

The future of IT infrastructure might well be based on open systems and commodity x86 architecture, but history has taught us that nothing stands still. Somewhere, someone, will be developing a new technology which has the power to turn everything on its head and start another new evolution line.  Until then, hyperconverged solutions are real and represent a major step towards simplifying the data centre. Anyone considering an infrastructure refresh should, at least, consider and evaluate a hyperconverged solution, because the benefits they can bring are substantial and the hype that surrounds them is fully justified.