Crushing edge complexity with automation

Just a few years ago, a power supply issue at a major airline data center led to a cascading series of problems that crashed its IT and business operations worldwide, canceling thousands of flights and stranding tens of thousands of passengers. It was a public relations disaster that in large part resulted from a relatively minor issue—re-plugging a power outlet—that could not be identified and managed quickly enough before its consequences spiraled out of control.

And despite the notoriety of the event and its obvious business repercussions, something just like it can still happen at any moment to any organization.

Traditional enterprise technology infrastructures are just too complex to manage. Despite decades of investment in tools and systems, the current IT management landscape is siloed and more complex than ever. In an Enterprise Strategy Group survey, more than 60 percent of IT decision-makers said their infrastructures were more complex than they were just two years earlier. And in another study, more than three-fourths of survey respondents said the complexity of their environments presented the biggest barrier to the productivity of their IT operations.

A new IT landscape

The enterprise technology environment cannot be managed effectively or efficiently without help. Across the board, the decentralization of network systems and compute resources has upended the game. The requirements of big data, artificial intelligence (AI) platforms, and edge architectures exceed the capabilities of data systems that were never designed to handle them. And all these trends are dramatically accelerating in the wake of the COVID-19 pandemic.

Overnight, enterprises challenged with managing a fragmented IT landscape find themselves in a world in which their entire workforce now works from home. That same phenomenon is driving a renewed urgency for accelerating digital transformation. But perhaps most challenging is the proliferation of new edge devices and infrastructures: Nearly 9 billion Internet-connected devices are in use today—and that number is projected to triple to nearly 56 billion by 2025.

Please read the report: Edge computing yields deeper insights, faster

Deploying IT workloads and supporting infrastructure across such a decentralized environment is a challenge that requires software management capabilities outside the typical enterprise data center. There are just too many devices to manage and too many issues that could occur, many of which cannot be understood. So complex and challenging is the situation that nearly 90 percent of business leaders surveyed said their organizations would need intelligent automation just to keep pace with changing business requirements.

The revolution will be automated

Automation offers businesses a way to gain control of digital transformation and the tsunami of new edge infrastructures. When automation combines with AI, the edge is transformed. It allows enterprises to take control and manage their edge infrastructures in much the same way they manage their on-prem data center and cloud operations.

So, what needs to be done to create the intelligent automated edge? There are three core components:

1. Holistic infrastructure intelligence and control
2. Integrated software, including third-party workloads
3. Infrastructure sensors and closed-loop remediation

The first core capability requires adding AI and machine learning (ML) techniques into the software that runs edge data applications and processing. Infusing the edge with intelligence allows it to operate on its own. It understands and reacts appropriately to unusual behavior or other user or operational deviations from the norm.

Please listen: How intelligent automation is transforming IT systems management

The second core component is management software that orchestrates seamlessly with third-party tools and workloads. Very few, if any, enterprise environments are greenfield, meaning any intelligent infrastructure requirements need to augment, not replace, existing infrastructure investments.

The third core component is providing the sensors that will connect, communicate, alert, control, and solve problems on the edge. This last piece is especially critical for edge environments in environmentally complicated locations.

These three components provide the critical capabilities for an intelligent automated management system that can improve performance from the network to edge computing infrastructure and software.

Automated edge management systems can see and anticipate problems on the edge before they begin; reallocate workloads in the event of a problem; and, using closed-loop remediation, repair and reestablish themselves independently. This frees enterprise IT staff to focus on new value-creating tools or services for the organization rather than monitoring screens for alerts.

Major benefits

So, what are some the benefits of automating network management operations?

• Reduced complexity: The typical enterprise has a large investment in legacy tools and systems. Managing such tools and systems was a complex process even before digital transformation and the rise of edge infrastructures. By unifying and automating these systems, organizations can be centrally managed from a single pane of glass. Multiple data streams can now be aggregated using AI and ML techniques, minimizing touch points and the amount of data needed to be humanly evaluated.
• Automated workload protection: Automating workload protection on the edge provides the eyes and support needed to identify issues before they become problems and automatically troubleshoot and remediate problems before they become crises. Automated workloads can reallocate workloads geographically so they continue to function without interruption.
• Lower downtime and costs: Automating edge infrastructure reduces network downtime and its associated costs. A Ponemon Institute study estimates the cost of data center downtime is $9,000 per minute, or $540,000 per hour. But just as valuable is the fact that edge systems are traditionally the biggest revenue producers for most companies. One survey projects that by the year 2025, automated edge systems could produce as much as $5 billion per year in new revenue.
• Digital transformation support: Automating edge network management can demonstrate the success of the organization's digital transformation initiatives and encourage buy-in at all levels of the enterprise. Successfully rolling out an automation edge management initiative is a powerful proof of concept for validating a digital transformation project and how to handle edge use case deployment at scale.
• Improved security: The FBI reported that since the start of the coronavirus pandemic, cyberattacks have increased 300 percent, with 95 percent of all enterprise security breaches due to human error. Especially now, in the new normal of work from home, remote workers continue to be the principal target of cybercriminals. Automating security from the network to the edge enables the enterprise to automatically enforce identity, device, and access security policies across all enterprise end points.

Given these benefits, here are some best practices leading enterprise adopters suggest for automating edge infrastructures faster and more efficiently:

• Architect for the future. Enterprises should begin with an overall strategy that looks beyond the initial use case to how automation will be implemented across the entire organization. Future-proof the infrastructure so it can adapt to new technologies and scale easily in response to changes in the business.
• Build in AI. Preventive maintenance capabilities in AI infrastructure are critical to creating an automated intelligent edge. Such functionality is key to successfully managing the complexity of the technology environment and especially for building in proactive and eventually self-healing capabilities across the entire IT environment. One expert recommends that organizations make this element a priority in planning for three-year technology infrastructure refreshes. Equally important is the value of AI and data. Collecting and analyzing data with AI is a key capability that will continue to grow in importance in the years ahead.
• Create a baseline. Many strategies are siloed, as successful automation planning requires many different organizations to be involved. It helps to view the enterprise holistically: What are the business units doing? What workflows are third party? Design for an end state that integrates all of these aspects of the organization.
• Focus on low-hanging fruit first. For many global organizations, the problem is often with their Wi-Fi systems or older switches. Go after that low-hanging fruit first. Demonstrating success with those initial use cases will result in multiple benefits, which is critical to continuing the momentum of digital transformation. It shows leadership that digital transformation is a net gain for the organization and deserves continued funding for the next use case. But just as important, the improved user experience generates momentum from the bottom up.
• Take a consultative approach. Enlist a partner with domain expertise. Many use cases cross multiple domains. Expertise in this area can greatly reduce the pain and speed the gain of the automation strategy and overall progress of an enterprise digital transformation.

The bottom line

Ultimately, the goal of an organization's automation strategy—the metric on which it will be judged a success—is how it changes the user experience for the better, or for a service provider, how it improves the customer experience. In almost all cases, that means making the edge experience the same as that in the data center and the cloud.

Enterprises need help to manage the complexity of their evolving edge to cloud environments. Automation is the answer.