Subscribe

Say what?

No really, what is hyper-convergence, in real terms?

Paul Ruinaard
By Paul Ruinaard, regional sales manager, sub-Saharan Africa at Nutanix.
Johannesburg, 28 Jul 2015

Contrary to what others try and sell, a hyper-converged solution is not just a box with everything on it; it is more, and less, than that. What it is, however, is possibly one of the most disruptive technologies facing IT and the data centre since virtualisation. Why? Because it truly does deliver on the notion of 'consumerisation of hardware'.

When flipping open the textbooks, hyper-convergence has been defined as such: a type of infrastructure system with a software-centric architecture that tightly integrates compute, storage, networking and virtualisation resources and other technologies from scratch in a commodity hardware box supported by a single vendor.

What is hyper-convergence?

That's great. But what does it really mean? Hyper-convergence lets users do more with less. It tightly integrates and places all traditional data centre hardware requirements on one box that the user can then manage centrally. It is easy to scale, easy to install, and even easier to replace. In order to expand a hyper-convergence-based infrastructure, all that is needed is to add nodes to the actual base unit.

The need for hyper-convergence was born out of the acknowledgement that legacy data centre infrastructure, as it now stands, no longer meets the high density and high performance computing demands of the modern data centre. The modern data centre is a resource-hungry animal that needs to scale rapidly in order to house the volumes of data. That said, the enterprise storage, servers, and networking infrastructure of yesterday simply aren't keeping up.

Today, IT needs to focus on five key deliverables, namely: service delivery, efficiency, centralised management, scale and competitive advantage. While cost still features extensively, the real pressure lies in the operational efficiencies IT can deliver and the speed at which these can be rolled out.

What is Web-scale IT?

The advent of hyper-converged computing solutions saw the growth of what is now termed 'Web-scale IT'. While the concept is seemingly new, its principle is founded on a process that dates back to the industrial revolution - the need to build a better production process.

Those techies who helped Facebook or Google to build their data centres - as their businesses and their need to house more data grew - will know that playing Tetris with the server rooms of the past to ensure they could fit on a smaller footprint was the inspiration behind hyper-convergence, and subsequently, Web-scale IT.

At the heart of hyper-convergence is automation and simplicity. Hyper-converged devices are able to automate and rationalise infrastructure in order to meet the needs of businesses with a voracious appetite for data. The key, however, when first developing these systems, was to ensure they would be as reliable and as stable as the infrastructure of yesterday - while at the same time meeting the demands of today's technologies.

So, what was needed was a way to enable cloud-centric economics and scale, coupled with the scale and performance served up by the bloated infrastructure in older (albeit stable and reliable) data centres.

Making hyper-convergence work

While building this new production process, a couple of key factors had to be taken into consideration. A hyper-converged solution needed to integrate all the qualities required to build a data centre infrastructure - processing power, networking and storage - in the quantities people would want them, and then offer so much more.

Central to the birth of hyper-convergence has been the reinvention of virtualisation for the age of the cloud. The new flavour of efficient virtualisation and the ease of use and deployment of the software has made life for IT that much easier. Today, spinning up servers is now measured in minutes, which leads to instant gratification, but hardware and infrastructure lags. However, thanks to Web-scale IT, the hardware side of Web-scale, namely hyper-convergence, can keep up.

Playing Tetris with the server rooms of the past to ensure they could fit on a smaller footprint was the inspiration behind hyper-convergence.

That said, true hyper-converged systems are able to offer myriad benefits and features, not least of which include: data efficiency; elasticity; VM-centricity; data protection; VM mobility; high availability; and critically, cost efficiency.

The data centre providers of today want to be able to deploy a data centre that enjoys the same ability to scale as those of Amazon or Google. That is one that isn't geographically large, but which houses enormous processing capacity, can scale up or down beyond traditional barriers, and of which the "hardware" component has become an easy-to-replace commodity.

All of these requirements are placing enormous strain on legacy IT infrastructure that to date remains siloed and is still unable to deliver on the scalability of the cloud. This is why hyper-convergence is considered as disruptive as it is - because it truly blows open the lid on how things have been done and how some companies still sell the data centre.

Hyper-convergence allows a business to take a long-term view of its investments, fuel growth and innovation, all the while improving efficiency and agility, and dramatically reducing the total cost of ownership.

Share