The use of private and public cloud is maturing; with that maturity, companies are evolving how they decide where to place workloads, which can include cloud repatriation.
When the practice of cloud repatriation began, it referred to the concept of moving workloads—otherwise known as applications—out of the public cloud to a private cloud alternative, including on-premises. With the introduction of industry concepts such as cloud-first and cloud-native, many companies may not be doing cloud repatriation in the true sense of the term but aiming for something in the middle.
According to a 451 Alliance Datacenters 2021 survey, 48% of over 600 respondents indicated that they had moved at least one workload away from the public cloud to another venue in the past 12 months.[1] This survey statistic begs the question: does moving a single workload away from the cloud equate to cloud repatriation, and if so, why? If not, is the practice of hybrid infrastructure becoming the new normal for selective cloud repatriation?
This first article of a series on the evolution of cloud repatriation will answer these questions and provide a roadmap to help companies navigate public and private cloud—including on-premises—with a hybrid infrastructure approach.
Trends driving the evolution of cloud repatriation and hybrid digital infrastructure
The term cloud repatriation implies that workloads moved out of the public cloud will permanently reside elsewhere. Herein lies the debate over whether repatriating workloads is an either-or scenario. Even cloud service providers are now offering solutions that extend to on-premises IT environments for customers choosing cloud repatriation—demonstrating that they understand some workloads are best run on-premises, but may still need to connect to the cloud.
While businesses may realize cost savings when moving workloads out of the public cloud, organizations seem to be making more of a philosophical shift when considering where to place their workloads. Flexibility is at the forefront of their minds. Concepts such as “best execution venue” for workloads speak to how organizations are no longer looking for a single all-encompassing solution to meet their IT needs, but rather an IT estate that offers flexibility and accommodates cost, performance, governance, and security. That’s when hybrid digital infrastructure can play a key role.
Ultimately, the cloud repatriation discussion centers on two important trends in enterprise IT: application mobility and hybrid architectures. According to the same 451 Research study referenced earlier, the movement of applications between environments is no longer a one-time event; instead, it’s becoming an ongoing IT consideration that illustrates how companies value flexibility. One might say the effect of moving applications to or from the cloud has become a revolving door, rather than a boomerang.
There are several factors driving the movement of workloads between public and private cloud, data centers and on-premises, including:
- Unsanctioned use of public cloud.
- Security concerns.
- Application lifecycle management.
- Regulatory/governance requirements.
- Data locality or sovereignty regulations.
Given the rising use of hybrid architectures, leading cloud service providers have introduced new solutions, such as AWS Outpost, Azure Stack, Google Anthos and IBM Satellite, for managing workloads in a blended approach of running applications on-premises and in the public cloud.
Optimizing environments and deploying hybrid architectures
Let’s talk optimization. Optimization is to hybrid infrastructure what lithium-ion batteries are to electric vehicles: it’s what makes the engine go! Deploying an architecture that offers the flexibility to move to and from hybrid environments easily is one way organizations are future-proofing their infrastructure investments. Just as the pay-as-you-go model within IT has allowed organizations to become nimbler, scaling quickly to meet changing capacity needs, administrators are finding applications need to be able to move depending on business factors. Defining the criteria for which workloads should run in the cloud and which should stay on-premises is a key first step.
Whether it’s initiatives to streamline cost, moving workloads for compliance and/or governance requirements, avoiding vendor lock-in, or introducing new infrastructure, workloads and their environments are not stagnant. They shift, move, and are set in the “best execution venue” for that period of time, based on what the business currently knows. That’s why it’s more important than ever to provision secure, fast, on-demand connections between your private and public clouds; essentially, taking a hybrid infrastructure approach. Local users can do this across the globe using Platform Equinix®.
Leveraging Platform Equinix to take a hybrid infrastructure approach
Depending upon their architectural needs, customers can leverage Platform Equinix to deploy virtual or physical infrastructure on-demand for a hybrid infrastructure approach. Take the example shown below. This diagram highlights the power of Equinix Fabric™ software-defined interconnection, which acts as the stitching mechanism that links colocation, Network Edge virtual network function (VNF) devices, and Equinix Metal™ on-demand automated compute, together with corporate networks and service providers. Each of the Equinix products shown below allows for low-latency on-demand connections that can be spun up or down as business needs change.
Architecture for On-demand Physical and Virtual Infrastructure
To learn more, read the Platform Equinix Vision Paper.
In our next article, we’ll talk about why taking a hybrid infrastructure approach can be an optimal solution for managing workloads across cloud, physical and virtual infrastructure. You’ll also discover how to design hybrid architectures that morph to accommodate your dynamic business needs.
[1] 451 Alliance, part of S&P Global Market Intelligence, ”Cloud Repatriation: Are Companies Moving Away From the Public Cloud?”. Pedro Schweizer, November 2021.
Subscribe to the Equinix Blog