Adaptive Applications Require Automation

Last fall, I wrote about the role of apps in adaptive applications. To refresh your memory and save you the trouble of re-reading it, let me sum up: apps are the focus of adaptive apps, specifically the goal of bringing intelligence and real-time changes to ensure digital services meet user and business expectations. Today I want to focus on automation and the notion of ‘real-time changes’ because our research shows we’ve got a long way to go to enable them.

Real-Time Changes Require Automation

To actually change operating conditions in real time will require the use of automation. This is because the process of scaling up – or down – or changing the primary location of an app for a given user to improve performance means changing configurations in multiple services.

In any given app deployment, there are already many services involved in the delivery and security of the application. There’s DNS, of course, and there’s local and global load balancing, ingress controllers, API gateways, and likely a WAF or API security service in addition to core network security like DDoS protection.

Any one of these services might require a configuration or policy change in order to adjust the real-time operating condition – availability, performance, and security – of an application. If we assume a mostly manual method of adjusting configurations and policies, well, we’re assuming it’s not going to be “real-time.”

So, we turn to automation. Data-driven automation, to be precise. The kind of data that’s generated by the entire stack supporting a digital service and analyzed by systems and AI to produce insights that can be turned into actions that lead to improvements in security, scale, or performance.

Many of these insights are as basic as "site is overwhelmed, redirect to a secondary site," and some of the insights may be more subtle, such as recognizing that a single app or component within the digital service is performing poorly due to load and determining that automatically scaling the component is the best way to improve performance (by reducing load, of course).

Regardless of the action suggested by an insight, a manual response is simply out of the question. By the time the right person is aware of the issue and executes the command to make the change, the user is likely long gone.

Thus, we assume that automation is the answer, at least for well-understood responses to problems.

The State of IT Automation

Unfortunately, for organizations seeking even the most basic real-time changes, the lack of automation in critical components of the stack is likely an obstacle. While our research shows the lower orders of the stack - mostly network, systems, and app infrastructure – are well on their way to being fully automated, the domain of app delivery and security is not.

Both domains remain startlingly low in terms of automation, which points to difficulties ahead for organizations seeking to enable the kind of real-time changes needed to enable adaptive applications.

automation-NWC.jpg

A deeper analysis indicates that this is a universal condition for organizations at every level of digital maturity. That is, even those organizations that exhibit maturity in most of the capabilities required to operate as a digital business still show a surprising immaturity in the area of app delivery and security automation. 

This is an area where organizations need to focus on if they’re going to sustain progress on their digital transformation journey toward becoming a digital business. As more business functions within IT are digitized (and our research shows that is happening), more applications will need to be operated, and we're all well aware of the Law of Diminishing Deployments: more operators is not the answer, automation is.

Now, there’s a lot more to this “automation” than just scripts and integration with ITSM. There’s also a significant amount of data that needs to be ingested, analyzed, and turned into insights. That’s why “observability” is also a part of an enterprise architecture for digital business. Observability feeds the automation necessary to make the real-time changes required to deliver adaptive applications that meet consumer and corporate expectations and deliver real value to the business.

Adaptive applications is the goal, but the path to get there goes through app delivery and security and a whole lot of automation.

Related articles: