Javascript on your browser is not enabled.


Pankaj Nath

Pankaj Nath, AVP & Practice Head - Cloud Services - Netmagic (An NTT Communications Company).

More From The Author >>

How Web-Scale IT Helps Rethink Enterprise Datacenter of Tomorrow

The new approach to data center infrastructure design and management that augments agility and cost effectiveness—which was primarily designed by and for service providers to provide web or cloud services—is turning out to be a lot more relevant to enterprise datacenters, whether managed in-house, co-located or on cloud.

Once it was famously said, there’s one thing common between a full grown tree in a jungle and the way IT ecosystem has evolved in an enterprise. Cut the tree and you can figure its age by counting the concentric rings in the trunk of the tree. Similarly, IT ecosystem traditionally has been growing incrementally, one layer over the other.

Now let's face it. If you've made a considerable investment in terms of time, effort and money to architect your IT ecosystem to grow incrementally and in structured scale up, you may end up discovering that some of it may not make it to the digital era. Unfortunately, your datacenter may prove incapable of efficiently coping with a sharp increase in demands of the digital business imperative. In comes Web-Scale IT that promises to radically rethink datacenter architecture to help an organization deal with massive scale.

The term Web-Scale IT, coined by Gartner, basically describes the new approach that most of the large cloud service providers and companies that offer services running on massive cloud ecosystems have adapted to deliver seamless user experience on a huge scale. This approach enables organizations with both scale and speed to shrink time to market for business services powered by IT functions and slash infrastructure costs while infusing agility, creating a conducive environment to stimulate IT culture change and augment the quality of service.

Simply put, Web-Scale IT is all about adopting the best practices learnt from cloud and IT-based service providers, such as Google, Netflix, Facebook etc., running super scales of software-defined datacenter architecture and automated IT ecosystem, and applying these principles and best practices at enterprises that operate in markets of high volume and rapid scale.

Getting Ready for the Next-Gen Growth

Speaking at the IDC Directions conference in San Jose last year, Vernon Turner, Senior Vice President of enterprise systems at IDC declared that by 2025, approximately 80 billion devices will be connected to the Internet. “Over 4,800 devices are being connected to the network as we speak. Ten years from now, the figure will escalate to 152,000 a minute,” he said.

This has opened up floodgates for information processing requirements. Organizations that inherently are dealing with the massive scale of data exchange and information collaboration will have significant ramifications on how their underlying datacenters and relevant infrastructure will handle such loads.

These organizations can be from a wide variety of verticals, ranging from operating in the space of e-commerce, to a cloud service provider, to manufacturing firms adopting machine-to-machine communication platform, to logistics enterprises upscaling to IoT ecosystem, or to IT services firm dealing with plethora of service requests from across the globe, or to stock exchanges running millions of transactions every minute.

When it comes face to face with the challenges of a data intensive age, the traditional IT infrastructure will crumble and the datacenters built around the time-honored architecture will find hard to scale up and out. The organization will increasingly have to adopt a modular, software-driven, flexible and open platform that can easily integrate and support the channels of cloud services, mobile computing, social networking, and big data processing to continue to run the user services irrespective of the scale in demand.

Web-Scale IT: The New Approach

Digital paradigm is expanding like the shadow of a cloud on the ground. Naturally, it has set the Web-Scale IT juggernaut rolling in the enterprises. According to Gartner, Web-scale IT is one of the technology trends most likely to have a significant impact on enterprises over the next three years. By 2017, Gartner had predicted that Web-Scale IT will be an architectural approach found operating in 50% of global enterprises, up from less than 10% in 2013.
"Web-scale IT looks to change the IT value chain in a systematic fashion.

Datacenters are designed with an industrial engineering perspective that looks for every opportunity to reduce cost and waste. This goes beyond redesigning facilities to be more energy efficient to also include the in-house design of key hardware components such as servers, storage, and networks. Web-oriented architectures allow developers to build very flexible and resilient systems that recover from failure more quickly. The approach includes open hardware and DevOps to increase datacenter agility and cost efficiency," said Cameron Haight, Research Vice President at Gartner.

The new approach to data center infrastructure design and management that augments agility and cost effectiveness—which was primarily designed by and for service providers to provide web or cloud services—is turning out to be a lot more relevant to enterprise datacenters, whether managed in-house, co-located or on cloud.

Web-Scale IT and Enterprises

By definition, Web-Scale IT is all about industrial data centers that have agile and collaborative processes running on web-oriented architectures with programmable management to provide unprecedented scalability, speed, and stability. In the Web-Scale IT architecture, the datacenter isn't viewed as a congregation of components, rather it is operated as a whole, with all data, metadata, and operations distributed across the entire cluster, while being put together across heterogeneous subsystems. This, interestingly, sets it apart from the hyper-converged architecture.

Certain enterprises and solution providers erroneously refer to Web-Scale IT and Hyper Converged architecture interchangeably. The primary difference between the textbook definitions of Web-Scale IT and hyper-converged computing environments lies in the way abstraction, software, hardware, and application layers are coupled or decoupled from one another. Unlike hyper-converged architecture's reliance on hardware abstraction (virtual disks), Web-Scale IT works with reliable software abstractions that allow distributed object and file system storage frameworks to offer seamless consistency and scalability.

Pure play Web-Scale titans such as Google and Facebook are more likely to use custom-built hardware and heavily customized applications that are tightly coupled with the hardware to manage unprecedented scale, unlike the hyper-converged environment's idea of having infrastructure software decoupled from the hardware, and application decoupled from the virtualization layer, all running on mass-produced infrastructure to reduce costs.

Now here lies a caveat. Just as systems designed to conventionally work on small-scale computing will break down when stressed with scale, ecosystems built for large-scale environment can’t naturally adapt to smaller fitment. Most of the enterprises are least likely to come close to the gargantuan scales of the Googles and Facebooks of the world. That's why the Web-Scale IT approach can't be lifted as is to be placed in an enterprise production environment. The Web-Scale IT version for enterprises rather combines the best of both the worlds of Web-Scale architecture and a hyper-converged environment.

By marrying the cloud and new technology enablers such as Software-Defined Network (SDN), Network Function Virtualization, open communications and more with integrated compute and storage to significantly improve performance and scalability, an enterprise can combat issues around vendor lock-in, shadow IT and siloed infrastructure while reliably dealing with unprecedented scale.

Undoubtedly, by applying the Web-Scale IT principles coupled with hyper-converged best practices, enterprise data centers would be better geared to take swifter, simpler, and more reliable steps to better handle the demands of the digital era.

Disclaimer: The views expressed in the article above are those of the authors' and do not necessarily represent or reflect the views of this publishing house

Tags assigned to this article:
Enterprise Datacenter

Around The World