Replatform vs Rebuild: Managing Legacy Apps

Businesses have long recognised public cloud as a solid foundation for a scalable, nimble IT infrastructure suitable for handling most business operations. Most have already undertaken cloud rationalisation projects to define which applications should move to Platform as a Service (PaaS). The line of business applications that could easily be rehosted or lifted and shifted have long since been migrated, leaving a handful of stubborn applications that, for a multitude of reasons, have yet to be cloud-enabled.

It never seems like the right time to address these applications. Quite frankly, they are hard work, needing extensive time and resource, and still carrying the possibility they may need to be rebuilt or replaced. So, these applications continue to sit outside of preferred platforms, monitoring and internal support processes while introducing risk into the business.

The trouble with rearchitecting

If an organisation needs to keep these applications, which were not designed for a cloud-first approach, then they may need to be rearchitected or rebuilt. The process of rearchitecting an application is complex and lengthy. Once the existing architecture is understood and mapped out, a full redesign to enable it to be hosted on PaaS can be carried out. At the same time developers need to get to work on the code; from basic cleaning by removing static content and server-specific references to recoding large elements and all while, ideally, leaving the UI/UX that end users are familiar with.

This approach carries excessive cost and requires extensive work on the server architecture even before rearchitecting the application software itself.

Replacing the application with one designed for the cloud could be considered but only as a last resort. While it removes the application from the transformation effort, it creates major problems for other business units who will have an entrenched user base and existing support models. The business will then need to test alternative new solutions and address business process, data migration, integration and training. However, in most cases, it is possible to avoid rebuilding or replacing an application completely by the application being replatformed.

Replatforming – the sweet spot

Replatforming is similar to rehosting, however, it involves some of refactoring for the purpose of taking advantage of the new cloud PaaS infrastructure. There are some common modifications that are usually performed during replatforming. For example, the IT team may choose to modify the way the app interacts with the database to benefit from automation and a more capable database infrastructure. This refactoring involves reworking the application to suit the new cloud environment and involves modifying the code to take advantage of cloud-based features and the extra flexibility that comes with them.

By replatforming, an organisation can enjoy the common characteristics of cloud including increased stability, reliability, easier user management and role management, as well as maintaining operational control and enabling DevOps to have common management and monitoring tools.

Many of the applications that have yet to move to the cloud are client-server based with the vast majority built upon one server to many clients, utilising the end user’s workstation and network for communication. The drive towards operational efficiency – delivering more with less – and increased mobility, combined with tighter compliance regulations means that is no longer a practical delivery method.
A simple way to replatform such applications could be achieved by delivering the application in a virtual desktop session using VMware Horizon, thus achieving the majority of the cloud benefits, such as centralised hosting on a preferred platform, streamlined support and mobility without the need to invest heavily in a full rebuild. Here’s a case study detailing our work with Turnkey Group that utilises this method.

Take applications on a case-by-case basis

If the main driver for moving applications to the cloud is to directly save money, then this cannot always be achieved by replatforming. The approach requires a range of skill sets and resource that organisations may not have in-house, which is often the reason why the line of business applications remain outside of preferred infrastructure and process.

Replatforming cannot be achieved without an investment of time and money, however, the outcome justifies the efforts. The cost of the project will compare favourably to the cost of completely rebuilding a solution or the wide-scale disruption if an application is completely replaced.

Dealing with your remaining line of business applications needs to be prioritised as they introduce risk into your business by falling outside of standard operational procedures. We believe a complete rebuild can usually be avoided by replatforming the application. At very least, it doesn’t hurt to explore this option.

The replatforming project will require resources to carry out the initial assessments, architect a design before carrying out any development and finally dealing with the migration. All of this requires end-to-end project management and a range of skills that many businesses do not have in-house.

Found yourself in this position? Cloudhelix can help; our extensive cloud hosting and consulting expertise will help you choose the right path and breathe a new lease of life into legacy apps. Read more about our application replatforming service here.