5 App Modernization Best Practices

An organization’s digital transformation is ineffectual if it doesn’t bring out the best from its applications. This is a near-universal concern since most enterprises still have core applications that are architected primarily to work in legacy IT landscapes in an on-premises dedicated environment, slowing down IT and business performance, stymieing innovation and increasing vulnerability to risk. Businesses are paying the price of their technical debt by forsaking agility, adaptability and competitiveness. That’s bad enough in regular times; in a post-pandemic scenario, this can have devastating consequences. 

Acquiring a few cloud-native applications that are not integrated with existing systems is of limited value and, at best, provides just part of the answer. On the other hand, overhauling the legacy application landscape is usually out of the question. The only reasonable solution, then, is application modernization. But where does an enterprise begin?

Pick and Shift

A great place to start is by examining existing applications. Understanding every application’s function, performance and issues and what it can gain from modernization will help to establish a ‘universal set’ to pick from. The candidates for modernization can be prioritized based on technology factors—for example, when the architecture adds cost, vulnerability or complexity—as well as business considerations, such as compromised agility or value potential. Finally, the enterprise needs to weigh the potential benefits, especially ROI, against the difficulty of modernization.  

Typically, the biggest, most important applications at the core of the enterprise are the hardest to modernize. But without their involvement, the transformation will most often be incomplete. Rather than lifting and shifting these monolithic applications and their vast data to the cloud, the enterprise can enclose them in an API layer to connect them to all the other applications that are being migrated as well as any new applications that are deployed. Using APIs to expose the functions and data of a legacy core system, such as ERP, allows it to be widely used, reused and modified, as well as leveraged to build new solutions and services. 

Split Into Microservices

Refactoring is a useful way of modernizing legacy applications that involves breaking them into microservices, which are then connected to cloud infrastructure such as Docker or Kubernetes. These microservices, which are loosely coupled, can be ‘called’ and scaled individually. Since refactoring requires significant rewriting and restructuring of the code, it is best done progressively over several iterations to make sure everything is functioning properly.  

Provide Full-Stack Visibility

Even when an application is modernized into microservices on the cloud, it needs to communicate with other legacy systems in the enterprise. The problem is that now there are more services and environments to monitor and secure, not to mention single points of failure, which not only increases complexity but also denies the enterprise full-stack visibility. But visibility across environments—edge to core (on-premises) to cloud—is imperative for success. This is achieved by ensuring that the modernized applications fit within the organization’s data fabric. The enterprise needs to capture data across the stack, from end to end, and process it in real-time to have an uninterrupted view of the performance and availability  of its applications.

Modernize the Process Landscape

A perfectly executed application modernization program can fail to deliver results if the surrounding requirements—for example, planning and meeting quality standards—are not met. Legacy processes can be a drag on modernization efforts. It is therefore necessary to address this issue by rearchitecting processes to fit a more automated setup by supporting code developers with teammates who understand the business requirements, adopt best practices and follow the example of successful digital natives. 

Track Performance Metrics

Going back to the point about visibility; enterprises need to establish key metrics to track so they can monitor the performance of applications right down to the code level. And it’s not just applications—the entire infrastructure, from cloud to databases to networks, needs to be tracked continuously to identify problems before they blow up. The performance of applications needs to be compared with the corresponding pre-modernization values to understand the benefits as well as identify areas needing optimization.

Balance and Act

Application modernization approaches range from rehosting and refactoring to rearchitecting and replacing, which are progressively costlier and riskier, but also progressively more impactful. Enterprises need to decide what’s best for them based on considerations of effort, cost, risk and result. While some enterprises may go slower than others, eventually they will have to modernize all their applications to be competitive and meet their customers’ needs in the most effective way.