Enterprises Are Starting to Decentralize Their Network Architectures

The steady rise of cloud computing has compelled many enterprises to pivot away from data center hyper-consolidation and toward more decentralized network architectures.

The opening months of the new year are always an interesting time for IT commentary. Some people are looking back on what happened during the past year — others are looking forward to what may happen in the year to come. But a thread common to both approaches is the conception of “progress” as a single, linear trendline that only moves in one direction: forward.

With exciting IT developments like quantum computingtwisted light, and a new WiFi standard on the horizon, one can hardly be blamed for focusing the bulk of their attention on the radically new — we’ve done plenty of that ourselves. But, on occasion, it’s equally worthwhile to examine subtler trends, trends that are often byproducts of the pendulum of common practice slowly reversing course. While these sorts of developments are certainly less flashy than major technological breakthroughs, they have the potential to be just as consequential in the long-run.

One such potentially game-changing industry trend is the shift from hyper-consolidated corporate data centers to more distributed, decentralized network architectures. During the early years of the decade, many large enterprises consolidated IT operations that once passed through 80 or 90 data centers into fewer than ten facilities. For a number of reasons, this practice is rapidly falling out of fashion.

More Clouds in the Forecast for 2018

The steady growth of corporate cloud computing is perhaps the most powerful driver behind this renewed interest in network decentralization. Once pigeonholed as a customer relationship management (CRM)-only approach, the last several years have seen enterprises turn to cloud computing for an increasingly wide variety of IT operations.

Indeed, according to RightScale’s 2017 State of the Cloud Report, an astounding 95% of large enterprises either have plans to run or already run some of their workloads in a cloud environment. But drillling down past this deceptive statistic into the kinds of clouds enterprises are using reveals an important clue as to why decentralization has become the name of the IT game.

Enterprises clearly harbor no desire to restrict themselves to a single cloud environment — in fact, enterprise cloud users run applications in an average of 1.8 public clouds and 2.3 private clouds, and all told, 85% of enterprises choose some sort of multi-cloud approach. Even more revealingly, while the share of enterprises relying exclusively on private clouds dropped slightly from 2016 to 2017, the share of enterprises relying exclusively on public clouds grew from 18% to 22% during the same interval. In other words, enterprises are becoming increasingly comfortable not only with hosting their workloads in a cloud environment, but in a cloud environment owned and operated by a third party whose hardware may very well be located hundreds or even thousands of miles away.

Edging Away from Traditional Network Architectures

That said, even as enterprises gravitate toward a cloud-first paradigm, the proliferation of Internet of Things (IoT) devices on corporate campuses will make it impossible not to keep at least some core infrastructure local. The further afield an enterprise’s servers are, the greater an issue network latency becomes, which can be hugely problematic for IoT systems that are dependent on constant data transfers between devices and management applications.

In addition to splitting one’s infrastructural portfolio between on- and off-premises networking hardware, one way enterprises can mitigate latency issues is by experimenting with edge computing. As its name suggests, edge computing entails assigning preferential treatment to time-sensitive workloads and processing them at the edge of a network architecture. Instead of requiring data-heavy, time-sensitive requests to travel back and forth across an entire network architecture, thereby increasing the risk of latency, edge computing allows these requests to be fulfilled as close in proximity to their originating source as possible.

Ultimately, no matter how an enterprise decides to distribute its architecture, proper network design and management is essential to optimal performance and robust cybersecurity. Fortunately, by partnering with a networking expert like Turn-key Technologies (TTI), an enterprise can rest assured that it’s getting the most out of its networks. In this day and age, a high-functioning network is the sail to the enterprise ship — without the former, the latter is going nowhere fast — and working with TTI guarantees strong tailwinds 24/7, 365 days a year.

By Chris Voll

02.13.2018

Sign up for the TTI Newsletter