Big Data Infrastructure Is About The Availability Of Services

High-performance computing data centers have been rapidly growing, and there’s a building need for a new set of benchmarks focused on big data infrastructure.  Sam Johnston, Director, Cloud IT services, Equinix, talked to SiliconAngle’s Founder John Furrier on the new trends of data center design and operation, and big data infrastructure in today’s demanding environment (full video below).

Appliance-like bundles of servers, storage, and software can show higher efficiencies for specific workloads than working separately as general-purpose pieces.  Johnston said data centers today are more homogenous in nature than heterogeneous. From building out special niches on the data center floor to bridging systems management tools, the industry is moving from a resource-based manufacturing system to a knowledge-based manufacturing system, where hardware, software and infrastructure all are working in conjunction. The adoption of modular, hardware and software enabled, data center technology will have more main stream solution for enterprises, governments and service providers.

In response to today’s market demand and where enterprises should move forward, whether mobile, cloud, or big data, Johnston says technology is rife with challenges and opportunities alike. For example, big data is growing exponentially and enterprises everywhere are being forced to grapple with storing, managing and extracting value from both structured and unstructured data. The technology needs to be re-engineered to handle vast amount of big data information. Big data solutions like NoSQL are being used not only by small enterprises but also being used heavily by Google and Facebook.

Hybrid IT is another area where IT is slowly moving. He said Hybrid IT is transforming IT architectures and the role of IT itself and IT leaders have started combining internal and external services, usually from a combination of internal and public clouds and data centers in support of common business goals.

When asked as how an IT organization should work, Johnston said, “rather than pulling a network to the data, IT should start pulling the data where the network is.” IT is now heading from Mainframe to client server and from client server architecture to cloud computing, mobility and virtualization. Multi tenancy, multi-tier data centers with virtualized applications are the need of the hour today and IT should start working in that direction.

On the managed services or outsourcing front, Johnston said IT organizations are taking an “adopt-and-go” strategy to satisfy customer IT consumerization requirements. He sees virtualization and cloud management platform will further evolve, where one can imagine migrating a virtual machine from an on-premise private cloud infrastructure to a public cloud provider like Amazon and vice versa.