I&O leaders will need self-organizing systems to handle digital complexity.
The goal of digital business is often to increase simplicity or efficiency for the end user or consumer. Yet for IT infrastructure and operations (I&O) leaders who build and maintain the systems that support digital business, complexity is increasing rapidly.
“The Internet of Things (IoT), edge computing, advanced analytics and the demand for real-time information and services will drive deployment of more complex and dynamic systems,” says George Weiss, vice president and distinguished analyst at Gartner. “The rapid growth of public cloud computing and hyperconverged infrastructure (HCI) highlight the approaches I&O leaders are using to meet these demands.”
IT leaders need to think bigger in scope and adopt principles of self-organizing systems of intelligence in data centers, edge and clouds of the future
One of the promises of these new approaches has been to simplify provisioning and integration by architecting IT infrastructure in a modular way to deliver services on demand. According to Weiss, however, a different reality is emerging that will require I&O leaders to evaluate systems and technologies in radically new ways.
“The increased simplicity of ‘as-a-service consumption’ and software-defined integration is more than countered by the complexity of an expanding IT cosmos comprising millions to billions of data points and gateways in a matrix of connections,” he says. “The application of artificial intelligence (AI) and machine learning (ML) to IT operations can mitigate the demands of this complexity but will require a shift in attention and focus.”
Silos Are Dead Ends
One potential roadblock to an intelligent, self-organizing data center is that many of today’s systems are implemented for specific-purpose workloads or business units, rather than broadly deployed and managed as a collective. This severely limits the potential for IT operations to continually optimize resources across different workloads. In turn, this will require AI and ML to be applied to IT operations.
“IT leaders need to think bigger in scope and adopt principles of self-organizing systems of intelligence in data centers, edge and clouds of the future,” says Weiss. “The goal should be to architect platforms and services that monitor and analyze system behaviors, resulting in continually optimized outcomes to predefined goals and service levels.”
Weiss sees four key properties of an intelligent, self-organizing data center:
- It can monitor, probe and provide feedback on configuration, workload, capacity and connectivity in traditional integrated and hyperconverged functions.
- It can monitor other systems in a mesh or cloud network that combines, composes and adapts to predetermined global objectives.
- It can deliver purposeful actions that fulfill the goals set by the organization.
- It can apply composable principles to resource utilization and efficiency.
With recent advances in algorithmic processing and ML, this vision is, in part, already within the technical capacity of some systems architected around HCI and integrated principles. Gartner therefore urges I&O leaders to map the foundations of next-generation self-optimizing data centers during the next three years.
To prepare effectively for this shift, I&O leaders should push the boundaries of their understanding beyond the capabilities that their systems currently possess, and seek out vendors that see such a vision within their own framework and objectives.
“Three areas of understanding around this trend are vital for I&O leaders,” says Weiss. “First, how it will benefit the business in terms of reliability, resilience, responsiveness and recovery. Second, how to effectively evaluate vendor deliverables, partner contributions and the wider business ecosystem. Finally, how existing legacy automation tools can be supplemented or replaced by the next wave of solutions.”
Gartner clients can learn more in the following report: "New Generations of Integrated Systems Will Apply AI and Emerge as Self-Organizing Systems of Intelligence," by George Weiss, et al.