How to create an AIOps strategy for the multicloud world

By Greg Adams, Regional Vice President, UK & Ireland, Dynatrace
AIOps are great for helping organisations to manage complex multicloud environments, but without the right strategy they can cause more damage than good

In an effort to increase agility and scalability, many organisations adopt multicloud environments and rearchitect their applications into microservices and containers. However, these cloud environments introduce new challenges. They are far more complex and dynamic, containing hundreds of technologies, millions of lines of code, and billions of dependencies. Additionally, many organisations still rely on DevOps teams to manually monitor these environments and handle issues that arise. In an ever-evolving cloud environment, the frequency and scale of change is simply beyond human capacity to manage.

To tackle these complexities, many organisations turn to artificial intelligence for operations. AIOps combines big data and machine learning to automate IT operations processes, such as event correlation, anomaly detection, and causality determination. While these capabilities can transform digital strategies, AIOps doesn’t fit every challenge. To develop an effective AIOps strategy, organisations must be clear about the problems they’re trying to solve and which AI approach provides the best solution.

Not all AIOps approaches are made for multicloud

Traditional AIOps approaches use machine learning to reduce alerts and simplify dashboards, so DevOps teams have less data to filter. This takes a statistical approach to AI, ingesting data to understand problems and make decisions based on past experiences. As such, it is well suited to automating many simple, routine, and easily repeatable tasks.

But before machine learning engines make any decisions, a data scientist must create an algorithm to train the AI and refine the technology by filtering out false positives. This process can take months or even years. Additionally, because it’s not autonomous, this type of AI is difficult and time-consuming to scale. What’s more, every significant change requires machine-learning-based AI to relearn the rules of its environment. This means AIOps that relies on machine learning is at its best in situations where processes don’t change often, such as legacy on-premises environments. 

Moreover, when it encounters a previously unseen problem, machine-learning-based AIOps is often unable to determine the root cause with precision, as it hasn’t learned the necessary rules. As a result, if a new anomaly in a large microservices environment triggers a storm of alerts, DevOps teams have to manually find the cause and try to catch the first domino before the pileup reaches an end user. Unsurprisingly, these challenges limit the scope of machine learning for tackling the complexity of today’s dynamic multicloud environments.

The next generation of AIOps is deterministic

Machine-learning’s shortcomings have led to the emergence of a new generation of modern AIOps that better supports today’s multicloud ecosystems. These approaches use deterministic AI, which performs a step-by-step fault-tree analysis based on a complete map of the multicloud environment. The environment is then continuously and automatically updated through real-time observability data. 

Deterministic AI provides precise answers about issues that arise. Because it has the full context that comes from its understanding of the organisation’s entire multicloud environment, it can suppress millions of unrelated events to trace the problem to its underlying cause. Deterministic AI can do so in near-real time, without requiring human assistance to learn new rules or analyse and interpret data. This AIOps approach can optimise the entire digital value chain by enabling DevOps teams to automate more complex, repeatable tasks that must instantly process a huge volume and variety of data. 

Embracing more intelligent operations

An effective AIOps strategy ultimately enables organisations to free up one of their most valuable resources — their developers. However, organisations must consider their approach carefully. Traditional AIOps based on machine-learning works for simpler tasks, but only modern approaches can truly keep up with today’s dynamic cloud environments. 

With a deterministic AIOps approach, DevOps teams are empowered to innovate. They are no longer bogged down by repetitive, manual tasks and can focus on more innovative aspects of software development, deployment, and delivery. This enables them to embrace more fulfilling, rewarding work and, ultimately, deliver more tangible benefits to the business.

Share

Featured Articles

Building Cyber Resilience into ‘OT in Manufacturing’ webinar

Join Acronis' webinar, Building Cyber Resilience into ‘OT in Manufacturing’, 21st September 2023

Google at 25: From a Search pioneer to AI breakthroughs

Technology Magazine explores how the tech giant went from being based in a California garage to a pioneer in technologies from AI to quantum computing

McKinsey: Nine actions for CIOs and CTOs to embrace gen AI

McKinsey identifies nine actions to help CIOs and CTOs create value, orchestrate technology and data, scale solutions, and manage risk for generative AI

OpenAI ChatGPT Enterprise tier drives digital transformation

AI & Machine Learning

Sustainability LIVE: A must-attend for technology leaders

Digital Transformation

VMware and NVIDIA to unlock generative AI for enterprises

AI & Machine Learning