Understanding your different R options
Despite all the talk about digital transformation and cloud migration, most organizations today still depend heavily on software and hardware in on-premises data centers. Recent data from Gartner shows that more than half of enterprise IT spending still goes toward traditional, on-premises infrastructure. But modernizing workloads and moving workloads to the cloud aren’t the same thing. Simply lifting and shifting an application as is to the cloud – rehosting – might be the easiest way to migrate, but it doesn’t let you make the most of the cloud’s many advanced capabilities. Worse, it might mean that you take all of your current software and application inefficiencies with you into the cloud.
You can reduce the risk of those types of inefficiencies by following a different R strategy: replatforming.
This involves modifying your applications before migration to take greater advantage of the cloud’s capabilities. For example, you might deploy a database into Amazon Relational Database Service (Amazon RDS), a collection of cloud-based managed services delivered as Platform as a Service (PaaS). This eliminates the need for database patching, which saves time on operations and improves availability and performance. In this way, you can quickly start to reap benefits from a move to the cloud.
Replatforming can be easier and more cost-effective than some other R strategies.
Another option is refactoring, which requires more time and effort than rehosting or replatforming. By rearchitecting some of your existing software code, you can optimize the outcome of a migration and gain big benefits from cloud-native services. For instance, moving to services like AWS Lambda, a serverless compute service, or to a managed service like Amazon API Gateway, can significantly reduce your infrastructure costs and provide greater scalability and fault tolerance. Consider this approach only when the investment required makes sense, though: you’re not likely to want to take this route, for example, for an on-premises application that’s not mission critical and works effectively and cost efficiently as is.