Bringing your Microsoft workloads to AWS: Why and how you might want to do it...

Microsoft workloads have been the backbone of organisations’ IT strategies for decades: 70% of all Enterprise Applications are Microsoft-based and over three quarters of those are still running on-premise. The technology in use is likely to include legacy .NET versions running on Windows Server and SQL Server databases licensed with an Enterprise Agreement, writes Rhys Jacob, CTO at AWS consultancy D55

But as cloud technology has evolved, and digital transformation continues to expand and reach new heights, Microsoft Workloads run on-premises aren’t fit for purpose. In fact, 48% of on-premise Microsoft workloads are forecasted to migrate in the next two years, making it one of the largest market opportunities in core IT and strategic focuses for CIOs and CTOs.

Businesses that remain running on these workloads on-prem face a range of challenges; the inability to scale their infrastructure for peaks in demand (unless they’re willing to spend a lot to do it), typically high operational and maintenance costs for their growing on-premise technology, inability to be agile without wholesale code changes, and not being able to digitally transform quick enough, impacting their ability to remain competitive.

Despite these barriers for growth, the question of whether to migrate Microsoft workloads, particularly onto non-Microsoft cloud services like AWS, can be a difficult one for CTOs to answer. If teams are thinking about migrating Microsoft Workloads to the cloud, Microsoft Azure might seem like the obvious choice – but in reality, AWS is running more Microsoft applications and at a better price.

While the thought of migrating all your Microsoft workloads onto the cloud may seem daunting, it has become such common practice that providers like AWS have streamlined the journey to the point where application downtime is non-existent.

Migrating Microsoft workloads to AWS: What it looks like.

The first step is understanding what needs to be migrated and what workloads can be retired is key. Once this is clear, applications can begin to be reallocated to EC2 Windows Servers and then ultimately to ECS/Lambda. This supports the lift and shift process, including SQL Server databases, which for most enterprises delivers immediate benefits, like reducing costs and increasing agility. The temptation for many here, however, is to pause the transformation journey once they have realised the immediate return on investment of a successful lift and shift.

As with any migration to the cloud, however, to gain the real benefits of it, including improved agility, speed to market, reduced cost, better scalability and reduced energy consumption, businesses must take a cloud native approach to their data systems. This means modernising and platform optimisation.

This can be done, to some extent, through re-platforming, whereby organisations migrate their applications without making wholesale changes to the architecture or code. This approach also means organisations can migrate their on-premise SQL Server databases to Amazon Relational Database Service (Amazon RDS). In doing so, businesses can continue using SQL Server but are no longer required to undertake time intensive tasks such as installation, patching, configuration and updates. However, the costly licensing fees attached to SQL Server will continue.

For the full benefits of cloud, re-factoring and re-platforming, which typically involves application changes and entire re-architecture strategies, is the step which allows organisations to truly untap the potential of cloud technology.

Here, businesses that no longer want to continue paying the licensing costs for SQL Server can move their database across to AWS Aurora, a fully managed database built for the cloud, and AWS Babelfish, which allows Aurora to understand queries from applications written for SQL Server, completing the database modernisation.

Meanwhile to re-platform a business’ applications, converting it to dotnet core and running the application on AWS’ Linux instead of Windows not only saves on the Windows Server license fee, but it also allows further re-architecture for greater modernisation, and supports organisations in becoming fully cloud-native. This means breaking down existing and legacy monoliths into more maintainable microservices, allowing each microservice to adapt and grow independently. Having a clean separation between microservices allows developers to focus on individual services without it impacting the broader system.

Crucially, microservices also allow applications to communicate with one another via. Put simply, when one application or service emits an event, other applications can be notified and decide whether or not they need to do anything with that data.

The benefits of modernisation are far reaching. From enhanced security to increased flexibility and lower licensing and consumption costs, organisations can unlock huge growth potential once their cloud infrastructure is optimised to support business objectives. It’s these benefits that have made 84% of AWS’s customers prioritise application modernisation in the next two years.

Final thoughts

Migrating Microsoft workloads to AWS may seem drastic, but the process to getting there has never been more streamlined thanks to AWS technology. In fact, AWS now has a 45% larger share of Microsoft workloads on the cloud than the next largest cloud provider, and it’s why our industry has reported a 23% increase in CXO level modernisation conversations. The process of migrating workloads is now becoming just a matter of time for most organisations.

Join peers following The Stack on LinkedIn