Accelerated over the past 5 years and further reinforced with the digitization of the economy and more recently with the COVID-19 crisis, migration to the cloud is still often opportunistic or even erratic. A holistic approach, by a global enterprise architecture methodology, brings consistency, controlled costs, and a long-term vision.
Migration to the cloud: various strategies applied
With the COVID-19 crisis, bringing applications outside of the enterprise is no longer optional. Today, digital links with the company’s ecosystem (prospects, customers, partners) are no longer sufficient and remote working is emerging as a new standard. Hence an ever-greater shift from on-premises to the cloud, and even to multi-cloud.
Applications can be brought to the cloud in different ways, according to varying degrees of optimization: Rehost (move the application to the cloud without rearchitecting it), Replatform (take advantage of some cloud optimizations offered by cloud providers, such as database-as-a-service), Repurchase (purchase an off-the-shelf SaaS application), and finally Refactor: the application is refactored - or re-architected - by leveraging cloud-native features.
These different cloud migration strategies each bring their own set of advantages and disadvantages, ranging from speed, immediate or longer-term costs, and even revision of the functional scope. In most organizations, these strategies have been or continue to be used opportunistically, depending on the applications involved and the ongoing timelines.
Enterprise architecture provides a technical and business context
To avoid making tactical or emergency choices, organizations should anticipate the way applications are moved to the cloud. Enterprise architecture is useful to understand the context and all the dependencies of an application for preparing and planning before the migration.
The enterprise architecture team will not only map the entire application landscape, but also will categorize and assign each application to the most appropriate migration strategy depending on their context and impact on business capabilities.
The main criteria are the value that each application provides to the business, their criticality, the way they are used, the technologies on which they are based, the data that they process, and their dependencies in the overall information system.
Depending on the strategy chosen for each application, different architecture skills will be called upon. While in most cases, business architects will intervene on the overall functional aspects of the application, solution architects will focus on applications to re-platform, i.e. re-architect using some cloud optimizations, or completely refactor them.
From opportunistic to automated holistic strategies
Whatever choices are made for each application, adopting a holistic strategy is always a good idea. It’s especially important to optimize associated cloud infrastructures which impact costs related to the migration of the information system to the cloud.
This comprehensive view also enables IT departments to get prepared and plan for future migrations. By refactoring applications – at least for the most critical ones – they can support future agility making the organization more competitive. Another benefit is the gain related to the seamless management of services in a multi-cloud environment enabling IT departments to move from one cloud provider to another.
In the long term and in regards with the complexity of cloud offerings, enterprise architects will be able to rely on automated recommendations based on artificial intelligence and machine learning. AI and machine learning can consider all the infrastructure optimization criteria for each application, including the service it provides to the business, its dependencies, hosting requirements, and governance aspects (regulatory).