Surely it must be a tough IT salesperson that can clinch a deal for so-called ‘infrastructure invisible’ right? It’s kind of like ‘cloud in at the deep end’ with converged compute (i.e. computer processing), virtualization and data storage intelligence all wrapped up in one bundle.
The shape of this bundle doesn’t even come in a nice shiny box — this is cloud, so it comes in a software-defined box located on a server box that exists in a datacenter somewhere that you may never see.
To make things even more intangible and abstracted, invisible infrastructure now comes with added ‘machine intelligence’ — and this means our computer systems will now start to work things out before you even we needed to worry about them.
As conceptually ridiculous as some of these ideas sound, this is pretty much how enterprise cloud infrastructure is starting to look i.e. invisible.
This is the gambit that Nutanix is tabling with its cloud platform software. The firm tries to warm things up by promising to deliver web-scale engineering that ships with ‘consumer-grade design’, but doesn’t that just mean it has an agreeable user interface and some nicely colored on-screen buttons?
The latest release of the firm’s core Acropolis 4.6 software offering has (yawn) more performance, but then… so does every major new release of enterprise software. More interesting perhaps is the freedom to choose hypervisors without lock-in and built-in machine learning capabilities that promise to increase operational efficiency.
Go ape for machine learning
This is not machine learning to ‘ape’ human intellect in the shape of Artificial Intelligence (AI), although it’s close. This is machine learning to automate our backend computer systems so that we don’t have to send the IT operations team in to sort out the guts of a problem when it occurs. Essentially the idea is that problems don’t occur; it’s all managed by software intelligence and analytics.
Nutanix promises the ‘operational simplicity’ of the public cloud without having to sacrifice the predictability, security and control of an on-premises infrastructure.
According to an official Nutanix statement, “Public cloud services provide a single runtime environment, which while suited to some applications is a poor fit for others. In the new era of enterprise clouds, flexibility and choice are fundamental tenets, allowing businesses to run traditional enterprise applications such as databases and VDI (Virtual Desktop Infrastructure) as well as next-generation big data and DevOps-[driven] applications uniformly and consistently on the best runtime environment.”
From many clicks to single-click… to zero-touch
The firm goes deeper on the whole invisible datacenter play by saying that enterprise clouds today require machine intelligence and automation to simplify complex operations. The idea is that we’re going from many clicks to a single click and eventually from a single click to zero-touch management. In this regard then, the new version release contains a product called Prism Pro, a datacenter automation solution with a search-first interface and customizable dashboards for IT operations.
“This release extends the ambitious Acropolis App Mobility Fabric (AMF) with critical software updates that give IT teams the freedom to pick the right virtualization environment for their applications and the flexibility to switch easily in order to reduce costs or simplify operations,” said the firm, in a press statement.
Prism Pro features something called X-Fit technology with foundational machine intelligence. This brings self-learning capabilities that continuously improve prediction accuracy and the quality of automated decisions over time. Not decisions about which apps are good, where to eat lunch and who to lay off next, but decisions relating to operations, such as smart capacity management.
The invisible future PC
So your future computer is invisible i.e. it’s just a screen interface (or tablet or mobile) with the backend in the cloud.
But more than this, your firm’s IT operations staff are invisible (or negligible at best) as your data pipe is run using machine intelligence with multiple predictive algorithms that play off against one another to pick the best fit and making actionable recommendations to eliminate over-provisioning.
All of this happens with the health of the applications themselves (and therefore the user) in mind.
If we were very futuristic we would even say that you should worry if you see computer systems in the office of the near future i.e. not seeing is believing that the cloud computing based software-driven services designed to serve you are working. For now though, still be careful of tripping over those wires from time to time.