Sophisticated migration of an IT infrastructure from on-premise data center to the AWS cloud

ConcertAI, one of the leading providers of medical research software,  has faced very common scenario for a mature healthcare data center as its infrastructure had developed over years to become extremely sophisticated but highly inconsistent. Some of the servers were running on an end of life OS and desperately needed upgrading. The client decided to implement a cloud-based solution for infrastructure migration and partnered with ABCloudz to make the system more manageable and prepared for the upcoming refinement. Our specialists suggested the customer to move to AWS because of its infrastructure control functionalities, high productivity, and the ability to delegate some maintenance tasks to the provider. Another important factor in favor of AWS was that some of the client’s infrastructure parts were already running on AWS.

We conducted a thorough technical analysis of ConcertAI’s infrastructure to come up with a comprehensive migration approach. As a result, we successfully migrated 144 client servers and 75 workstations to AWS EC2. This significantly improved the infrastructure’s manageability and prepared it for further modernization.

Read this post to learn how we solved the database migration challenge for the client.

Building IT Inventory

The client runs a large data center involving roughly 200 servers and 100 workstations. Some of their servers were already running on AWS. Meanwhile, the client’s infrastructure included SQL Servers (2000, 2005, 2008, 2012, 2014), web servers, email servers, and physical on-premises servers. Inconsistencies between legacy and less outdated servers made their system hardly manageable.

To make things worse, there was a severe lack of consistency in operating systems between these servers. Some of them were running on Windows Server 2003, various versions of Windows Server 2008, different versions of Windows Server 2012, Windows Server 2016, and Windows Server 2019. With such a varied infrastructure, the client faced critical problems with their system’s manageability and productivity.

Check out the diagram below for the information on the proportion of operating systems of the client’s servers.

One of the immediate needs that required attention was the servers with the older “End of life” operating systems. These servers needed an upgrade as soon as possible because Windows had no longer offered maintenance support. The client did not want to retire the outdated servers because there was some uncertainty about how this would impact the existing server connections.

With these restrictions in mind, it was critical to research the infrastructure and migrate it to AWS without any impacts to the data. The client agreed that an efficient lift-and-shift migration of their 144 inconsistent servers and 75 workstations connected to them was a vital step before proceeding with further system modernization.

Reverse Engineering of Data Flows

The client chose ABCloudz as a reliable partner with an outstanding cloud migration experience deeply rooted into the legacy of DB Best, an original company that had spent more than 20 years migrating databases, applications, and infrastructure. Our strong background in technology and extensive experience in database migration were the critical points that attracted the client to us.

As there was not enough documentation related to the client’s infrastructure and the client’s DB management specialist could not provide us with the full picture, we had to start with a thorough analysis of ConcertAI’s system. Our team applied custom tools to evaluate the overall state of the infrastructure and its configurations and determine the most efficient migration plan.

As a result, we identified a new challenge: the client had run a very sophisticated scheme of the server connections. Specifically, a single server could potentially be responsible for the functions from three different processes like data reception, processing, and analysis. Also, many servers were linked together with ETL jobs. Therefore, we had to be very cautious to avoid any disruptions in the client’s server connections.

Six waves of lift-and-shift migration

After thoroughly analyzing and identifying all the connections in the client’s infrastructure, we decided to divide them into separate groups according to their functionality and connectivity.

We delegated servers into 5 Virtual Private Clouds (VPCs). To ensure top-notch security of HIPAA-covered servers working with Private Health Information (PHI), we delegated these servers into isolated networks with established restrictions in communication with other servers.
The image below illustrates the design of the implemented VPC architecture.

To structure the migration and avoid any potential disruptions in connectivity, we divided the migration of 144 servers and 75 workstations into six waves. You may observe them in the image below.

The entire migration process took less than 3 months. For the transition period, we created a hybrid infrastructure and network configuration to ensure that the client’s data center kept working.

We migrated around 95% of ConcertAI’s servers with CloudEndure because the client had used it before and was comfortable with this technology. CloudEndure replicated partitions from source to destination and then switched between them.

However, CloudEndure was not suitable to move larger systems. Therefore, we had to use DataSync for the remaining 5% of the client’s servers that were too big for CloudEndure. DataSync was installed like a VM appliance in the on-premises datacenter and replicated over a secondary data link to AWS.

An infrastructure that runs smoothly without data loss

We efficiently migrated ConcertAI’s infrastructure to AWS EC2 without any system disruptions and data loss. The most notable benefits of the migration were:

• The infrastructure gained a sufficient level of consistency. The client no longer had to apply time-consuming physical scaling. Also, advanced AWS control functionalities enabled the client to manage their infrastructure more efficiently and within one cloud system. Such innovations improved the infrastructure’s manageability by at least 25%.
• The client was able to delegate some infrastructure management tasks to AWS. This enabled them to reduce the staff required for system maintenance. Along with the additional financial benefits from the AWS cloud (such as the absence of downtime fees), this action brought the customer significant cost reductions.
• ConcertAI received a uniform and consistent infrastructure. All connections between the customer’s servers were identified and preserved. Therefore, the system was ready for further modernization.

How can you benefit from our expertise?

Insightful? The most important thing is that we can tailor our analysis and migration techniques mentioned in this article to the needs of any client. Whether you want to modernize your infrastructure, database, or application, we’re ready to help. As proven experts in database migration, we apply our custom tools and approaches to solve the most difficult challenges. Contact us to discuss your database migration plans and take full advantage of system modernization!