The traditionally cautious data center market is on course for a major shift in focus over the coming 12 months, with implications for both the technologies and products enterprise CIOs will be looking to embrace.
At one level it will be business as normal for the data center in 2017, with continued virtualization of workloads, legacy integration and the seemingly, endless quest for greater performance and capacity. Beyond that, however, increasing acceptance of the public cloud as a key component of digital transformation will present hard pressed IT departments with a completely new set of challenges. As will growth in Big Data, the Internet of Things, process automation and other technologies increasingly seen as essential for the digital business.
Because data matters
One of the biggest shifts in 2017 will be a move away from a focus on computational power, to addressing the storage side of the equation, largely driven by Big Data applications and the tidal wave of data analytic tools that began to flood the data center in 2016.
On the hardware front, this will accelerate the already brisk march towards flash domination over conventional magnetic storage, with enterprises adopting all-flash architectures to both improve performance and eliminate silos. 2017 will also see vendors bring a number of disruptive new technologies to market, such as NVMe and 3D XPoint, able to deliver data at DRAM-like speeds but at flash-like prices.
These developments will take a while to hit the mainstream, but promise to significantly improve storage performance and agility while driving down operational costs. This will empower enterprises to look for imaginative new storage use cases, to help drive the business forward. It will also assist in the development of next generation storage solutions to meet a growing need to distribute storage resources, putting data as close as possible to where it’s needed without loss of performance, control or security.
One of the casualties of this shift will be the once dominant Storage Area Network, due to the growing availability of more scalable, alternatives without the management overheads of the SAN approach. Increasing availability, and a widening understanding of the benefits of data analytics within the data center, are also at work here and we can expect to see new tools to better balance both computing and storage workloads and automate the management of that process across a mix of platforms.
The end of the data center?
The mix of platforms that enterprise IT will need to embrace will increasingly include the public cloud. For some, this signals the end of the corporate data center as we know it, as the complexity of building and managing an on-premise infrastructure outpaces the ability of the IT department to contain it. However, the enterprise is unlikely to give up the data center without a fight and there are clear signs that a hybrid approach will not only prevail throughout 2017, but become the de facto standard for enterprise IT.
Vendors are coming to recognize this trend, even Amazon which once maintained that no one needed an on-premise data center at all, but has since changed that stance and made it easier to migrate and balance workloads across public and private domains. Converged hardware vendors are integrating public cloud technologies into their products with the aim of enabling enterprises to take full advantage of the instant scalability and pay-as-you-use economics of the public cloud, but without relinquishing the ownership, control and security afforded by on-premise alternatives.
Towards utility computing
And lastly, there are clear signs of steps being taken to tackle the inherent complexity of the hybrid approach, both at the infrastructure, application and business level.
On the infrastructure side, the virtualization technologies so successfully applied to the management of compute resources are now being extended to do the same for storage and networking. Products and technologies are maturing rapidly, with 2017 set to be the year that the software-defined data center (SDDC) finally becomes a reality.
There have been great strides too when it comes to managing application workloads and, by leveraging data analytics, the development of sophisticated business-level automation tools. Heralding, perhaps, the start of the much anticipated “utility” era of enterprise IT, these tools will enable organizations to define in pure business terms the policies by which they want their digital systems to be monitored, managed and measured. More than that, they can then be executed by smart software regardless of the platforms involved - whether in the on-premise data center, on externally hosted private clouds or public cloud services - to deliver the IT the businesses want, wherever and whenever they need it.
Paul Phillips is senior regional director, UK, Ireland & South Africa at Nutanix