Virtualization Technology News and Information
Pivot3 2019 Predictions: Data Center Technologies Continue to Become More Accessible

Industry executives and experts share their predictions for 2019.  Read them in this 11th annual series exclusive.

Contributed by Bruce Milne, Chief Marketing Officer, Pivot3

2019: Data Center Technologies Continue to Become More Accessible

People who try to predict certainties in information technology usually end up wrong (ask anyone who thought that the blockchain would dominate all industries in 2018). Instead of talking about certainties, it's better to talk about attitudes. While certain technologies may not have been widely deployed in 2018 , their patterns of adoption all reflected a certain attitude towards infrastructure agility and the importance of intelligence and automation. In 2019, those trends will continue, along with an increased focused on edge computing, as organizations seek to reap the benefits from an ever-expanding Internet of Things and the need for processing analytics and data in real time. We'll also see organizations come to terms with mixing public cloud, private cloud, and traditional data into a hybrid cloud that meets their changing needs.

As IT increasingly feels the pinch to deliver more services and support more mixed application workloads that are secure, seamless, and streamlined, composable infrastructure, edge computing, multi-cloud strategies, and artificial intelligence and machine learning will be driving digital transformation in 2019.

Composable Infrastructure Leads to More Efficient Operations

The endgame of composable infrastructure is the ability to run nearly any kind of workload on any kind of server or endpoint, with the necessary configurations happening instantly and automatically. The adoption of this kind of technology will have unexpected ramifications - for example, airports.

A real-world example is an airport customer, which wanted to add more airlines and more flights. Usually that means physically building more gates and terminals, but the airport didn't have room for that kind of expansion. Instead, leaders there realized that a lot of their gates and terminals stood empty throughout most of the day. They could increase their utilization by allowing more than one airline to use the same terminal-and this is where composable infrastructure began to come into play.

Software, as it happens, is the main barrier preventing multiple airlines from using the same terminal. Each airline uses a different set of applications for everything from checking boarding passes to ordering airline meals. Using composable infrastructure, however, the airport can rapidly re-provision the airline infrastructure at each terminal or gate, allowing different airlines to use each gate in rotation throughout the day.

As time goes on, composable infrastructure will allow the physical infrastructures of the world to become more and more flexible. Airlines are just the first example.

The Edge Will Become a Peninsula Instead of an Island

Right now, edge computing is an island - a silo, really. Because of bandwidth issues and the limitations of physical infrastructure, it is difficult to transfer workloads between on-premises, the edge, and the cloud. Therefore, many efforts in the core must also be replicated in the edge-instead of having a single application that distributes data throughout the cloud, core, and edge, administrators must often instantiate three separate applications in each location, all doing the same thing.

The coming year is going to bring the state of the art to a level where the network edge starts to become federated. This will happen gradually, but soon many companies will be able to manage the core, edge and cloud from a single pane of glass. This unlocks numerous possibilities, including the ability to automatically transfer workloads to areas where they can do the most good while using the fewest resources. Federating the edge means freeing up compute and storage throughout the entire network and adds a level of intelligence that allows for faster decision making and eliminates the unnecessary movement of data, ultimately lowering costs for the organization. This will all be made possible by combining advancements in machine learning with evolving hyperscale infrastructures which allow for faster data processing and provide better value to the business.

AI Will Continue to Evolve

Industries haven't yet hit a ceiling on the capabilities of artificial intelligence. In fact, artificial intelligence will largely enable the innovations above. Artificial intelligence will handle the task-scheduling requirements of composable infrastructure and the federation requirements of edge computing. Human administrators have neither the time or the ability to sense how to optimize a given workload. Only artificial intelligence can manage complex cloud-edge-core networks at scale.

Artificial intelligence can do this because machines are good at sensing patterns that humans can't, and the data center is full of patterns. Someday soon, an AI controller will be able to sense that a stack of servers is overheating, indicating that it's at maximum capacity. The controller will know that if it redistributes the workload, the temperature will return to normal. It will do that without ever alerting an administrator. In the near future, companies will increasingly use AI for automated error handling.

Hyperconvergence Becomes Specific

Hyperconverged infrastructure is becoming steadily less generic. By this we mean that the core technology-COTS servers, hypervisor, and VMs-remain the same, but the applications running on it have become steadily more specialized. It used to be that airports, banks, colleges and transit systems would all consume similar systems with similar configurations. Not anymore.

For the organizations using this infrastructure, specialization means being able to start using hyperconvergence immediately. Previously, an organization would purchase a generic infrastructure and configure it to their liking. Vendors are now onto this and offer specialized configurations that organizations can purchase a la carte.

In other words, hyperconvergence is now becoming mainstream-consumer grade, even, as are all the technologies above. Every advancement that makes hyperconvergence, AI, edge computing, and composable infrastructure easier to use also pushes that technology further into the public sphere. It's an encouraging sign that these technologies won't just make lives easier for systems administrators, they'll soon make our daily lives easier as well.


About the Author


Bruce Milne, Chief Marketing Officer, Pivot3

Bruce Milne is a visionary technology executive and Chief Marketing Officer at Pivot3, where he leads the company's growth in emerging markets that are ripe for transformation with IoT and smart technologies. With more than 20 years of experience, Bruce has played an instrumental role in establishing a vision and executing a go-to-market strategy for software companies that include Socialware, Hyperformix, and OpenText (formerly Vignette).

Published Wednesday, January 09, 2019 7:28 AM by David Marshall
There are no comments for this post.
To post a comment, you must be a registered user. Registration is free and easy! Sign up now!
<January 2019>