AWS Lift and Shift

The Challenge

Our customer wanted to modernise its strictly on-premises application - which supports remote cardiac monitoring and data reporting - to the cloud to better support future growth and technology rollouts.

Specifically, the company wanted to be better prepared to quickly leverage the latest and most advanced technologies with fault-tolerant and highly available cloud services and technologies.

The company was looking for a highly collaborative cloud partner to direct and guide a lift-and-shift application migration process and leave behind clear operational artifacts for the company to use afterwards.

Why Amazon Web Services

Our customer chose AWS as their public cloud provider for several reasons. The team understood that AWS is a robust and mature market leader that provides a scalable, reliable, and secure global cloud computing infrastructure. Most of the team was familiar with AWS, and their IT team had implemented some AWS products already. Overall, the technology team found the AWS Application Migration Service (AWS MGN) very straightforward and easy to use.

The Benefits

Their core application is a mission-critical system. The application needed to cut over to AWS in a timely and effective fashion. The team was eager to take advantage of significant improvements in cloud-based failover, load-balancing and availability technologies. Our customer was also keen to leverage Terraform Infrastructure as Code (IaC) to efficiently manage an ever-growing catalogue of cloud services.

There was no time wasted digging in. The team used AWS MGN to create an initial image from the on-premises environment and then used that image to create a template that was modified iteratively for the final system.

Multiple technical goals were delivered on, such as migrating SQL Server databases to Amazon EC2, which is more robust and fault tolerant. JBoss and TIBCO servers were migrated smoothly.

The TIBCO servers, in particular, were migrated with membership, ownership and file permissions intact. For JBoss servers, AWS MGN duplicated a single instance of each node, then created an image and three servers to support the JBoss user interface (UI), Dev interfaces, and the NEXUS UI.

Terraform was then used in a straightforward fashion to efficiently reuse images that had already been spun up.

Unique and unexpected migration challenges were encountered and effectively resolved with key stakeholders at each stage of the project.

For example, some environments simply failed to stge. In another instance, two app environments without access to public endpoints hampered AWS MGN functionality. AWS MGN encrypts all data, but the service relies upon access to public endpoints to migrate data.

For the application’s use case - specific to a local configuration – working collaboratively with with the customer and advanced AWS support resources to craft a workaround.

A specialised VPC was created that did not need to connect to public endpoints. The VPC leveraged conditional forwarders and AWS MGN agents were configured to use that VPC to connect to an on-premises service, ingest the data and move it privately. In another case, adding nodes to a failover cluster required close collaboration with high-level support resources at both Microsoft and AWS, together partnering closely to work with cloud providers to resolve the issues.

A robust, straightforward and repeatable IaC base was delivered that can be used against any environment - such as test, dev, qa, uat, prod - or with other products. Just as important as technology, the level of partnership was extremely collaborative and transparent.

Not only is IaC as a technology naturally more visible, but rationale and methodologies was explained at every step, making the why and how of the engagement much more obvious. The approach helped the customer progress towards strong best practices as the project advanced from test to dev and beyond.

The customer worked closely and iteratively with our teams to move through their configuration. Environments were delivered that became more robust, predictable and consistent as they progress towards prod. A logically hierarchical, consistent, and plain (requiring few calls out to libraries) code base was created. In any environment, variables, providers and main files are used in the same ways.

A decade’s worth of accumulated technical debt was resolved. Once the project got underway, the true volume and complexity of data became apparent. About 10TB of data representing millions of files and folders was migrated, mostly from production servers with just a single instance.

Although the engagement was envisioned as strictly lift and shift - no refactoring involved - efficient modifications have already been made to code that support special requests such as database changes and load balancing.

Note:

The work described in this engagement was originally completed by 1Strategy, a TEKsystems Global Services company. As of June 2023, 1Strategy has fully integrated with TEKsystems Global Services to continue to deliver AWS expertise to customers. Learn more about our AWS solutions.