Cloud repatriation may not be the first item that comes to mind when taking an inventory of the digital transformation toolkit for an enterprise, but it’s gaining traction as an option in certain cases where migrating data or assets away from the cloud turns out to be an advantage. As the term itself becomes more familiar to transformation teams, many still struggle to understand exactly what cloud repatriation means, when and where it’s best applied in the enterprise, and how to craft the right cloud repatriation approach for a particular organization.
Why repatriate?
Cloud repatriation refers to the process of bringing an application or service out of the cloud and returning it to an on-premises location, or in certain cases migrated to a SaaS host or outsourced to an MSP service. This approach is usually adopted to improve performance of the application or service or to reduce the costs associated with IT infrastructure and operations.
These scenarios are far from hypothetical, with industry analysts reporting a particularly strong increase in cases since the pandemic where organizations pursued hurried lift-and-shift migrations and later needed to revisit and restrategize those investments. Examples of cloud repatriation include undoing poor migrations where duplicate or misaligned infrastructures – such as a partially migrated client-server application, or multiple versions of databases placed on-prem and in the cloud – may be causing cost or performance issues.
In other cases, cloud repatriation may be a strategic response to new or changing business conditions. There may be a need to adjust to shifting storage needs, internal policies, licensing, or regulatory rules. Or perhaps a unified data management strategy or some other enterprise-wide project is prompting a need for interoperability that some vendors can’t support, requiring the transformation team to move those assets somewhere else.
Planning the cloud repatriation journey
The above use cases illustrate how cloud repatriation can be a useful and strategic option as the organization looks to modernize and transform. By the same token, their intricacies point to how complex an undertaking the cloud repatriation effort can be. The challenges of repatriation are playing out against the backdrop of modern IT ecosystems where cloud-native DevOps and other advances have created more dependencies and copious, multi-dimensional data flows.
As a starting point, cloud repatriation project teams have their pick of enterprise tools on the market – including from HPE, Amazon, and other companies – designed to stand up hyper-converged network, storage, or compute resources on-premise. But while these “cloud in a box” options are useful for establishing infrastructure and bundling up some of the complexities, organizations still face the burden of ensuring overall coordination – typically with some form of management plane that continually aligns and configures all assets, systems, and processes to work together.
All this complexity must be managed and accounted for in a cloud repatriation effort that happens against the backdrop of workload-intensive mixtures of legacy and modern applications and infrastructure. This is fueling massive increases in the volume, variety, and velocity of data to be managed. And enterprise IT teams may also struggle with outdated tools and operational processes that stifle their ability to respond to this spiraling complexity and keep up with evolving business and user expectations.
In this sense, the real challenge of cloud repatriation is the Day 2 reality that comes after the technology investments have been made – where multiple tools or systems may be provisioned and running, but not necessarily configured to deliver optimal performance without service interruption, data loss or security gaps. Adding to the challenge is the increasingly complex dependencies of all these assets across vast distributed architectures and multi-cloud environments.
Key priorities to guide the successful cloud repatriation effort
There's no single blueprint or roadmap for how to design and implement a cloud repatriation effort. Nonetheless, it is possible to isolate a few key priorities that can keep an organization on the path to success. Chief among these is that going back on-prem doesn’t mean going back to “doing things the old way” with the same old management or operational paradigms.
This priority underscores how, in a certain sense, cloud repatriation is a misnomer. When done right, repatriation is more a question of refactoring and getting optimal business value no matter where assets are located or operating. What makes this possible is the management plane we mentioned earlier – ideally, a control and governance framework that continually configures and optimizes all assets, systems, and processes to work together seamlessly.
visibility. Successful cloud repatriation requires visibility into how assets and systems are running on an ongoing basis – including when new infrastructure or systems are provisioned and added, and anytime configuration changes happen to switches, firewalls, load balancers, or other components. Such visibility is necessary to establish and maintain a unified service view of how the repatriated systems and assets behave across the entire IT ecosystem.
A final priority to mention in the cloud repatriation effort is proactivity, and this is where AIOps platforms can overlay predictive and auto-resolution capabilities onto repatriated systems by applying a rich set of analytical techniques and contextualized intelligence to proactively change, modify, heal, and optimize how they run. This can power robust decision support and predictive recommendations that accelerate both root-cause analysis (RCA) and accelerate mean time to repair (MTTR).
Conclusion: reaping the benefits of successful cloud repatriation
Cloud repatriation is gaining traction as a transformation strategy in cases where it makes business sense to move certain technologies out of the cloud. When done correctly, the successful cloud repatriation effort can contribute to a more agile, data aware IT ecosystem – one that aligns multiple applications and infrastructure layers, spanning multiple data centers and clouds, to build a common data model to solve a broad set of business challenges. A fresh assessment of the enterprise IT environment may show that cloud repatriation is in order, and the right approach to the effort will ensure maximum efficacy, cost efficiency and ROI for the organization.
