Toward The Self-Driving Data Center

Like self-driving cars, the data center that runs itself, manages itself and calls for help when needed is not far away. Even complex IT infrastructure that is notoriously difficult to upgrade and maintain is being automated, converged, made into building blocks or stacks, and managed via software, not hardware.

Visionaries have suggested that the self-driving data center is as inevitable as the self-driving vehicle, as IT staffs admit that machines can do most anything better than a human, and start putting the machines to work. Doing so enables agility – that most essential IT building block – so that leaders can respond to a changing business universe. Even for those “humanists” who disagree that machines can perform IT manager tasks better, the efficiency gained from offloading repetitive functions, or making connections between often unrecognized, disparate events, frees organizations to serve customers at a higher level.

Similar to vehicles, data centers are well along their march toward full self-driving capability, and a continuum of automation and analytics-based solutions is in place to save time, hassle and costs. Using the storage industry as an example, the continuum of new devices, virtualization, alerting, and orchestration technologies enable successively greater machine direction of resources, and less overt IT staff involvement.

Six key technologies are evolving that most challenging IT domain – data storage – further along the continuum of more fully autonomous operations. They are turning IT managers into business agility agents whose work enables their organizations to achieve higher aims.

(Image: Timofeev Vladimir/Shutterstock)

High-capacity flash

New storage media represent the first step toward self-driving storage, removing the IT manager from intensive management of the storage resource. Modern flash drives last longer and have a more predictable lifetime than traditional rotating media, thanks to wear-leveling and other advanced controller techniques. This translates to less management overhead tied up in replacing failed drives and keeping inventories of spares – and thus less demand for IT management time.

As new generations of NAND-based and phase-change SSDs deliver stronger price/performance and increased competition brings price parity to hard-disk drives (HDDs), analysts project that the transition toward flash will accelerate.

(Image: jules2000/Shutterstock)

Containers

Technologies like Docker allow applications and processes to be easily and quickly launched and terminated. This, in turns, makes auto-scaling a breeze, instantly and automatically matching application resources with demand, without the need for advance planning. 

(Image: Pexels/Pixabay)
 

Tiering and caching

Tiering and caching technologies both make it so that storage administrators no longer have to determine the optimal placement of data based on access patterns. Instead, storage systems use sophisticated algorithms to determine where data should reside at any point in time. As the internet of things (IoT) drives more data, tiering and caching become particularly critical.

(Image: markusspiske/Pixabay)

Software-defined everything

Software-defined everything (SDx) encompassing storage, networking, and compute, is fostering open and more interoperable environments. The SDx wave has steered the industry away from decades of lock-in. This creates platforms with cost-effective hardware upon which value-added, software-based services are built by the ecosystem, including automation of almost every aspect of IT. 

(Image: geralt/Pixabay)

Remote management

Recent startups have designed their systems from the ground up to be operated as a service. These systems are imbued with autonomy, resilience, and self-healing, to allow them to operate continuously with little or no intervention. Moreover, such systems are monitored by small and highly automated teams, and thus efficiency translates to reduced cost and complexity for customers. 

(Image: geralt/Pixabay)

Intelligent automation

Hyper-scale cloud provider service chains at AWS, Microsoft Azure, and Google Cloud Platform already are fully instrumented and integrated with highly automated server, networking, and storage management software. They cede control of many common IT functions in order to achieve scale, agility and a lower cost-structure. These cloud providers  do not yet perform all performance and high- availability driving and monitoring functions 24x7, 365 days year. However, as they continually expand automated IT functions in the search for competitive advantage, the more favorable their scale-out costs and performance will be.

(Image: insspirito/Pixabay)