Getting ahead of your data migration journey 

MJ Scholtz, Data Engineer at PBT Group 

Moving your organisation’s data to the cloud can be a formidable task. It has the potential to be complex, time consuming, and expensive. It is therefore not surprising that many companies decide to wait before migrating. However, once the move is complete it will bring about improved scalability, agility, and security. 

Of course, a data migration journey requires precision planning from the business and its implementation teams. Perhaps more crucial, it needs the buy-in from leadership to accept the culture shift essential to be successful. 

At a high-level, migrating to the cloud is a four-step process: 

• Step 1 – Planning 
• Step 2 – Choosing your cloud environment 

• Step 3 – Migrate data and applications 
• Step 4 – Testing and validation 

Method in the madness 

There are several proven methodologies to aid in the data migration to the cloud. 

Re-hosting 

Also known as the ‘lift and shift’ method. This involves replicating the existing organisational infrastructure onto the cloud. It is something that should be considered by small and medium businesses who do not yet have any long-term scaling strategies in place. Even though this is the quickest method of migration, it can also be expensive. 

Re-platforming 

Also known as ‘the move and improve’ strategy. It entails making small improvements during the cloud migration process. While scalability is an important factor in this process, the basic architecture of the application(s) remains the same. This method should be considered by enterprises looking to scale up their services and performance through cloud-based technologies, even though this method does not make full use of cloud capabilities. 

Re-purchasing 

In many cases, legacy software and architecture can become redundant, expensive to expand or maintain, and leave the company vulnerable to security risks. In such instances, it is critical to move to a modernised cloud solution. However, if end-users are not properly informed and trained, preferably prior to implementation, this migration will not be successful. It is also important to do extensive research to identify the solution that is best suited to the business requirements while accounting for the infrastructure already in place. 

Re-factoring 

This method takes a ‘back to the drawing board’ approach where the entire infrastructure is redesigned from the ground up. It allows the business to take complete advantage of all the benefits of cloud computing such as auto-scaling and serverless computing. Because this is a complete redesign approach, it can become quite expensive and time-consuming. Again, proper research must be done to ensure the solution is selected with the business requirements in mind and can provide the most value for the business. 

I have always recommended that a company consult with a trusted partner to help navigate the business through this process. If the implementation is done optimally and efficiently, it can return vast dividends in the long run. 

Retirement 

This involves retiring all outdated and redundant applications. If there is no value to be gained by migrating an application to the cloud it can either be downgraded or terminated. This helps to shift the focus onto the essential apps and services while reducing expenses. Furthermore, the ‘retirement’ methodology can increase overall security by getting rid of unnecessary modules. 

Retention 

In some cases, certain modules may not (yet) be supported on a cloud-based infrastructure and the data and information for these modules cannot be transferred as part of the migration process. In these instances, maintaining the modules on-premises can make financial and operational sense. 

Join me next time, as I discuss what happens once the data migration tasks have been completed and some of the considerations around managing the business data in the cloud.