olly - Fotolia

Tip

What is the self-driving data center and how does it work?

Self-driven data centers require little intervention to handle daily operations. Explore their inner workings, benefits, and what HPE and VMware are doing to make them a reality.

As enterprise workloads become more complex, so, too, do the IT infrastructures that support them. Many organizations are adopting technologies -- such as automation, software-defined resources and AI -- to help ease this burden. Not only do these simplify operations and improve resource utilization, they provide the foundation for a new technological movement in IT administration: the self-driving data center.

The self-driving data center requires little human intervention to handle day-to-day operations. Tools that support the data center work together to automate and intelligently manage infrastructure, all while optimizing operations and proactively addressing potential issues.

At this point, the self-driving data center is more a theoretical goal than a reality, though. However, with vendors like Hewlett Packard Enterprise (HPE) with Synergy and Composable Rack and VMware with vRealize AI (formerly, Project Magma), and advancements in IT technologies paving the way, don't be surprised if you see true self-driven data center capabilities in the not-too-distant future.

The benefits of a self-driving data center

A self-driving data center can offer a number of benefits, efficiency among the most important. It automates and streamlines operations that require IT resources, eliminating repetitive manual tasks, reducing service disruptions, and saving time and money.

Because the self-driving infrastructure incorporates intelligence, the data center benefits from real-time analytics that lead to actionable insights and faster outcomes. At the same time, automation and software-defined capabilities enable more streamlined management and quicker resolutions to problems.

Because the self-driving infrastructure incorporates intelligence, the data center benefits from real-time analytics that lead to actionable insights and faster outcomes.

This efficiency also frees up IT personnel for other efforts. Rather than spending their days making hundreds of decisions and carrying out routine tasks, they can focus on learning new technologies, acquiring more advanced skills and developing cutting-edge products. IT professionals spend less time provisioning, upgrading, putting out fires and performing other tasks, and more time moving an organization forward to provide a competitive edge.

Another benefit of the software-defined capabilities of the self-driven data center is a reduction in the types of errors that manual repetitive tasks can introduce. When operations are automated and backed by intelligent systems, operational accuracy becomes the norm. The more manual and repetitive the operation, the greater the chance of introducing errors, especially in fluctuating environments.

The foundations of the self-driving data center

In the traditional data center, IT personnel perform most operations manually, whether general maintenance, troubleshooting problems or resolving issues. But as a data center grows more complex, the more difficult it becomes to keep equipment and applications running smoothly. Virtualization helps to some degree -- by better utilizing resources -- but it still calls for a fair amount of manual effort.

To address these challenges, some have turned to systems such as converged or hyper-converged infrastructures. These can help simplify and streamline operations, but they still require a lot of handholding, especially the DIY systems. For example, administrators managing hyper-converged workloads might still have to address unexpected contention issues or disruptions in service.

One thing hyper-convergence has done, however, is rally the forces behind the idea of the software-defined infrastructure (SDI), which brings IT systems under the control of software, helping to simplify operations and better utilize resources. In the early days of hyper-converged infrastructure, the SDI approach applied primarily to compute and storage resources, but today's hyper-converged systems also include software-defined networking to create a more complete software-defined data center.

With SDI, a system not only becomes more manageable, it also lends itself to automation, supporting such development methodologies as infrastructure as code. Together with SDI, automation simplifies such processes as provisioning resources, applying updates, monitoring components, balancing allocations, taking corrective steps and carrying out a variety of other tasks -- without (or, at least, very little) manual input required.

infrastructure as code (IaC)
Software-based application delivery model infrastructure as code -- a step on the way to the ideal of the self-driving data center -- automatically provisions and configures infrastructure resources as needed to host and run applications.

But automation is only as effective as the controls in place to ensure that operations work as expected or, better still, continue to improve. This is where AI-based intelligence enters the picture. Today's intelligent systems incorporate AI technologies such as machine learning and deep learning, along with predictive analytics, to fully utilize the capabilities of an automated SDI.

The AI-based service collects system metrics from across monitored components, aggregates and analyzes the accumulated data, and uses those findings to determine what operations to carry out and the best way to do so. Intelligent services don't stop there. They continue to grow smarter as data continues to be collected and the environment becomes better understood.

Moving toward the self-driving data center

So, with each passing day, infrastructures become more automated, intelligent and software-defined. As systems become more extensive and sophisticated, we move closer to the true self-driving data center. We're already getting hints of this from various vendor products and services. HPE's two composable infrastructure products, Synergy and Composable Rack, embrace the self-driving model.

HPE Primera storage
Support for HPE Primera storage integration with InfoSight AI and machine learning analytics, brings AI-driven intelligence to the company's composable infrastructure products.

AI-based InfoSight -- HPE's intelligent service for monitoring and analyzing its systems -- supports both systems. InfoSight previously monitored only compute resources for Synergy and Composable Rack, but because the two composable infrastructure products now support HPE's Primera storage platform, which also falls under the InfoSight umbrella, they feature intelligence for storage resource now, as well.

VMware has also been hard at work on the automated and intelligent data center front with vRealize AI, billed as the first installment of the company's self-driving data center vision. VRealize AI is an intelligent SaaS platform that uses AI technologies to support automated self-tuning to make VMware systems more efficient and performant.

The new service is currently in tech preview. But VMware promises that once released, the vRealize AI engine will run, manage and tune performance all by itself.

Self-driving capabilities in IT infrastructures are already a reality. As systems become more intelligent, automated and software-driven, we'll continue to move closer to the goal of the self-driving data center. IT infrastructures will be self-diagnostic and self-healing, helping to streamline operations, reduce human error and free up IT resources to focus on more innovative opportunities.

Clearly, the self-driving data center is the way of the future. It's only a matter of time before we get there.

Dig Deeper on Converged infrastructure management

SearchWindowsServer
Cloud Computing
Storage
Sustainability
and ESG
Close