Despite increasing cloud adoption and digital transformation initiatives, up to 80 percent of organizations’ data still resides in legacy systems, often running on IBM Power midrange server hardware in the data center. Since this data is extremely important for making well-informed business decisions (especially if we’re talking about an ERP system or long-running key business application), many organizations face the daunting task of incorporating this data into data modernization efforts.

Let's look at how to handle this

Lift and shift legacy workloads

Until recently, many legacy workloads could not be moved to the cloud without rewriting application code because the IBM Power network architecture is incompatible with the architecture used by the three major cloud providers. But now there are Infrastructure as a Service (IaaS) options that provide the virtual infrastructure that allows Power logical partitions (LPARs), as well as traditional x86 (non-Power) workloads to run in the same environment without any refactoring required. These options allow IT teams to move IBM I (AS/400), AIX, Power-based Linux operating systems, or traditional legacy x86-64 bit workloads to the cloud without rewriting, rearchitecting, or re-platforming. This significantly reduces the risk of these key applications breaking by accident or the migration project from spiraling out of control. This often makes the project viable for IT teams that would otherwise consider it too risky, costly, or time-consuming.

For organizations running legacy systems on-premises, the data modernization process involves two steps: (1) migrate key workloads to the cloud via a solution such as Skytap on Azure and (2) connect these workloads to cloud-native analytics tools. While it’s possible to connect on-prem workloads to a data lake in the cloud, the extreme latency and cost involved in sending high volumes of data back and forth for every big data query make this option less appealing to most organizations. Plus, having legacy workloads in the cloud opens up many other opportunities for modernization or taking advantage of other cloud-native value-added services, so the data modernization project can create other secondary and tertiary benefits down the line.

This incremental approach allows legacy systems to move to the cloud unchanged, improving application and data reliability, performance, and scalability. Then, different pieces of the applications can be replaced with cloud-native equivalents, or other cloud services can be applied gradually, one piece at a time, in a controlled manner. Here's a link to a Microsoft TechCommunity post where I cover some of the benefits of leveraging a SaaS such as Skytap on Azure and Azure-native services to modernize your legacy data. And for more details on the “Strangler Fig” method used to modernize your workloads, check out this Blueprint for migrating on-premises applications to the cloud.

Open the door for data insights

After the legacy systems are up and running in your public cloud of choice, they can be connected to cloud-based analytics tools. This unlocks insights from previously siloed legacy data, allowing your organization to include critical legacy data in advanced analytics, data visualization, and AI/machine learning projects.

For example, if you’re working in Azure you can use Azure Synapse to extract data from a DB2 database to an Azure Data Lake, and from there run big data analytics queries against it. Data latency will be low and big data queries will complete very quickly since everything is running in the same cloud (it will likely be in the same region, perhaps even the same physical building). Azure Purview can then be used to provide data governance and compliance and Microsoft Power BI can be used to create visualizations based on the data. All of this leads to better, more data-driven business decisions for the organization overall. To see this whole process in action, check out this demo: https://vimeo.com/749748081.

Access to legacy data is necessary for organizations to realize the full benefits of data modernization and migrating those workloads to the cloud is a much easier and more viable step on the journey than it was until recently.