In the last few years, many companies that initially decided to implement cloud-first or cloud-only strategies in public cloud platforms are reconsidering their data strategies.
An article by Denodo CTO Alberto Pan states that “By 2026, over 80% of organizations building centralized cloud data warehouses or data lakehouse architectures will decide to migrate certain workloads to other environments, including alternative data processing systems within the same cloud provider, systems in other clouds, or even on-premises environments.”
The solution applied in those cases is called “data repatriation,” and it involves moving data and/or workloads from the public cloud back to on-premises or private data centers. Repatriating data has become a hot topic for organizations looking to optimize costs, improve performance, and regain control over their own data. But what are the main drivers for this trend, and what are its implications? In this post, I will explore this trend, explain why some companies are moving data back “home,” and how a logical approach to data management, enabled by data virtualization, represents a solution for a smooth repatriation process. In fact, a virtualized approach can be useful not only to resolve the need to repatriate the data but also to define a solid data strategy from day one, to avoid moving data back and forth.
The Ins and Outs of Data Repatriation
Data repatriation requires the partial or complete transfer of data and/or workloads from a public cloud environment (such as Amazon Web Services (AWS), Microsoft Azure, or Google Cloud) back to on-premises or private cloud infrastructure. Repatriation doesn’t necessarily mean abandoning cloud; on the contrary, it reinforces a data strategy that balances the multiple benefits of public cloud with the specific requirements of a corporate data management vision. Organizations can leverage the best advantages of each cloud solution, available on the market, for the optimal mix of capabilities.
What are the benefits of this strategy?
- Cost Efficiency: Cloud services are normally agile and scalable. However, their cost also grows when they are used at scale. To avoid that, it is key to plan and estimate the volumes of data and the complexity of the data flows, to avoid skyrocketing costs. When data is repatriated to on-premises or private clouds, not only are operations more economical, but better yet, companies can forecast costs to avoid unpleasant surprises.
- Regulatory Compliance and Data Sovereignty: Companies need to maintain compliance with diverse data privacy regulations that are also updated frequently. Depending on the nature of their data, companies may be required to store it within certain geographical boundaries or management processes, but this is not always possible with public cloud systems. Bringing the data back on premises enables companies to control where the data is stored, to maintain compliance in these situations.
- Performance and Latency: Performance is critical in certain sectors or applications, and from that perspective, the geographical location of data and applications is important. Does it make sense to run applications on-premises while keeping the data in the cloud? The answer to this question is a clear “That depends,” but in general, these kind of data flows will be severely impacted by network latency.
According to some analysts, the data repatriation trend will rise as companies aim for more hybrid or multi-cloud architectures, and they try to get the best out of these offerings while maintaining in-house control over critical data. The concept of data gravity also applies here, in the movement of applications and data flows closer to the data sources involved in the use case, which should mitigate the performance and latency issues listed above.
Seamless Repatriation
A logical approach to data management can play a significant role in data repatriation by reducing the risks associated with skyrocketing costs and data governance. In fact, a logical data infrastructure can help prevent the need to repatriate data, as it stores data, by design, in the optimal location.
Using the Denodo Platform, a logical data management solution, for example, it is possible to define a semantic layer to represent a “facade” of abstraction between the data sources and the final users/applications, so that all migration, repatriation, or change in the source or storage technology is completely hidden from the final users, avoiding disruptions in service. In this way, the Denodo Platform helps to relocate data, preventing the problems associated with the need for repatriation.
The Denodo Platform provides a full data management solution, capable of acting as a query orchestrator that can deliver data in real time, applying a variety of optimizations on-the-fly to get what the customer requested in the most efficient possible way while also enabling, through its logical architecture, proper access control and data governance.
- Denodo not only enables the seamless migration of data and workloads without impacting consumers but it also incorporates advanced query optimization and acceleration techniques that effectively reduce cloud costs—without relocating the data. Effectively controlling costs requires a comprehensive understanding of how each data source is utilized, as well as who is accessing it and how often, all of which can be easily accomplished with the Denodo FinOps Dashboard.
By integrating these features with Agora, the Denodo Cloud Service, it is possible to apply the concept of data gravity on top and position Denodo strategically near the primary data silos, to reduce network latency while improving response time.
Denodo acts as a query orchestrator for business user requests, determining who can access data, when, and how those requests are executed. This optimizes performance and cost efficiency while protecting sensitive data by eliminating the need to move it off-premises. These capabilities make Denodo an ideal solution for dealing with complex scenarios in which data repatriation can be a strategic decision.
