The stigma of cloud lift-and-shift
The “lift and shift” is one of many approaches for migrating workloads to a cloud service provider such as Amazon Web Services or Microsoft Azure. Specifically, the lift and shift moves the legacy workload and data to the cloud in a relatively unchanged fashion. In other words, the original service is moved to an IaaS cloud without being redesigned for cloud native operation.
Cloud purists frown upon the lift-and-shift paradigm because legacy architectures will rarely take full advantage of cloud capabilities. For example: a legacy service may have application logic and SQL server services baked into a monolithic virtual machine. A more cloud native approach would be to bifurcate application logic to microservices running on containers or Lambda Functions, and punt the database services to a PaaS like RDS.
The paralysis of cloud perfection
Redesigning apps for the cloud sounds ideal at first. However, I’ve observed clients take a year or more to decide how to re-architect legacy applications. While companies attempt to design architectural masterpieces for a cloud-native nirvana, the clock keeps ticking on their legacy data center; the place where systems lack chargeback models, geo redundancy, and elastic scale.
Here’s the problem: all of this “design” is based on somewhat generic best practice. Enterprises need experience in the cloud to understand the nuances and limitations of specific cloud components in order to create realistic target cloud architectures. The catch-22 here is that you must first be in the cloud in order to gain cloud knowledge.
Just get started
Your journey to IaaS excellence starts with a single step; and likely a clumsy one. Move an auxiliary service first, such as a virtual remote desktop or an encrypted backup archive. Choose a component that’s simple, relatively self-encapsulated, and has few dependencies.
Initial cloud application migrations are never perfect, yet IT teams will learn an immense body of knowledge during the migration process. Institutional cloud knowledge really is the key to unlocking cloud acceleration, so getting started sooner rather than later is crucial.
