Story image

Self-driving cars are cool, but what about the self-driving data centre?

11 Jun 2018

Article written by Pure Storage A/NZ regional vice president Mike Sakalas

Technologies such as artificial intelligence and machine learning are signalling massive shifts in how enterprises around the globe act on their business’ insights. All of these developments create and require large amounts of very fast data, which means we must put our data centres in the driving seat in new and bigger ways than ever before.

The term ‘data deluge’ no longer encapsulates the scale of what we’re facing. To prepare for the self-driving era, we must get closer to realising what I like to call the ‘self-driving data centre’. With every new advance across IT automation, converged and software-defined infrastructure, containers and VMs, we move towards a powerful technology future that keeps pace with expanding data needs.

Take self-driving cars, for example, which are going to churn through an absolutely staggering amount of data. A Mercedes AMG Petronas Formula One car already has hundreds of sensors that transmit a whopping 12,000 data feeds with every lap. When autonomous cars must navigate unpredictable environments beyond the race track, and make split-second decisions, the implications are far more profound.

If businesses are going to take advantage of disruptive new technologies, it’s become increasingly clear that we’re going to need a data centre that eliminates as many manual operations as possible. The self-driving data centre uses data to become smarter every day, is upgraded, managed, and maintained by software, and is inherently simple and predictable.

Much like self-driving cars, self-driving data centres are really about four key things:

Simplifying and offloading to automation

Elimination of manual operations is a key tenet of the self-driving future of our IT systems, particularly as infrastructures continue to grow in size and complexity. Simplification and automation have enormous roles to play, and we’re well on our way to fully realising the benefits across our data centres. The more we automate day-to-day data centre tasks, the more we can maximise operational predictability, and enable teams to shift from fighting fires, to working on projects that actually provide business value. Our first step in the mission towards the self-driving data centre is to remove the knobs and dials, and eliminate as much administrative pain as possible.

Collecting data and interacting with surroundings

Self-driving cars leverage cameras and sensors not only for guidance and navigation purposes, but also for safety. When it comes to the data centre, every piece of actionable intelligence is equally valuable. Sensing and being able to model the data centre environment is critical to understanding an evolving IT landscape, and collecting data in this way will be fundamental to becoming more predictive about how our data centres operate.

Learning from data to make better decisions and course-correct where needed

The data centre is an increasingly complex interaction of systems. As we move forward, data will help us answer some of our most challenging (and valuable) questions about workloads, performance, and more. Machine learning will further help us gain a more complete and comprehensive grasp on operations and how to best optimise for efficiency. We’ve only reached the tip of the iceberg here, but the implications for data centre uptime and performance are huge.

Creating next-level efficiencies and optimisation through network intelligence

Whatever you may call it, the main idea behind the self-driving data centre is really twofold: it’s about gaining an unparalleled ability to see exactly what’s happening throughout your infrastructure, as well as trying to predict future outcomes based on data. Network intelligence is crucial to effective data centre monitoring and management, and we’re working our way towards a framework that will better predict capacity, performance, failures, and more.

The self-driving data centre will give IT the tools and time needed to respond to a shifting business landscape and better serve customer needs. A blend of automation, analytics, and innovative infrastructure is getting us closer to a place where we can save our IT teams business-critical time and costs, as well as work off key predictions and insights about workload performance and data needs.

The self-driving data centre is self-managing, self-optimising, and self-restoring. I may be a bit of a geek, but that sounds much cooler than self-driving cars to me.

Google doubles down on hybrid cloud strategy
CSP is a platform that aims to simplify building, running, and managing services both on-premise and in the cloud.
In ongoing cloud war, Google to acquire data migration specialist
Google is currently behind AWS and Microsoft in the cloud battle, and it would seem this play is an attempt to claw some ground back.
Interview: CyrusOne’s new Europe president on aggressive expansion
In this exclusive interview Tesh Durvasula shares how the company plans to have a Europe data centre portfolio providing nearly 250 MW by the year’s end.
Enterprise cloud deployments being exploited by cybercriminals
A new report has revealed a concerning number of enterprises still believe security is the responsibility of the cloud service provider.
Pure Storage expands enterprise data management solutions
It has integrated StorReduce technologies for a cloud-native back up platform, and expanded its data fabric solution for cloud-based applications.
HPE launches new real-time processing edge platform
The platform is said to help communication service providers (CSPs) to capitalize on data-intensive, low-latency services for media delivery, connected mobility, and smart cities.
‘Digital twins’ entering mainstream use sooner than expected
The term ‘digital twin’ may sound foreign to some, but Gartner says it is rapidly becoming established among modern organisations.
Infinera launches new ‘disruptive’ network architecture
The new end-to-end network architecture is said to enable instantly scalable, self-optimizing networks that adapt to the demands of specific users and applications.