Cloud Repatriation & On-Premises Migrations

The cloud is the future — or is it? While statistics show that the public cloud continues to grow, a small but loud group is proudly going in the other direction.

David Heinemeier Hansson, the CTO of 37signals — the company behind Basecamp and HEY, among others — recently posted his controversial take on the subject, announcing that the organization would be leaving the cloud.

Hansson isn’t the only one, as moving data off the cloud has become quite the hot topic across the industry.

Why is cloud repatriation picking up steam in 2023? Let’s dig into what it is and reasons you might want to move back to on-prem.

What is cloud repatriation?

Cloud repatriation is the process of moving data, workloads, and applications that were previously moved to the cloud back to an on-premises data center or from a public cloud to a private cloud. It’s also sometimes called cloud reversal or de-clouding.

Reasons for cloud repatriation

With all the benefits the public cloud offers businesses of all sizes, why are some choosing to move some or even all of their data back on-premises? The most common reasons that make organizations regret moving to the cloud include:

Let’s look at each of these reasons.

The high cost of public cloud services

One of the main reasons that companies choose cloud repatriation is the high cost of maintaining and managing workloads. Many companies are struggling with cloud costs: 82% of respondents in a recent survey said managing cloud spend was their top challenge, beating out security and lack of resources. While the cloud can offer cost savings for some businesses, others find cloud costs increase over time due to many factors, such as workload complexity, high amounts of data transfer, and the cost of cloud services.

Many businesses also find hidden fees after switching to the cloud. In fact, 82% of IT decision-makers said they had observed unnecessary cloud costs in one study. From data transfer fees to licensing fees to storage costs and more, it’s challenging for companies to anticipate all the charges ahead of time.

(Check out the latest in cloud cost trends.)

Data security and compliance issues

Some organizations move their data back on-premises because they are concerned about security or regulatory compliance. In fact, 40% of respondents in one survey said that security and compliance was the primary reason for repatriating their workloads.

Although cloud providers offer robust security measures, certain organizations—like healthcare or finance — need more control over their data than a cloud environment may provide.

Greater control and flexibility

Some companies choose to repatriate their cloud services to have more control and flexibility regarding IT infrastructure. Some aspects include data privacy, control over performance, and data governance, which some companies find they can more effectively manage on-premises.

Data locality needs

Data’s physical location can be vital for organizations, especially operating regions with stringent data residency regulations. Transferring data back from the cloud ensures that it’s stored in a specific location and complies with local laws. For example, many organizations want to ensure EU citizens’ data remains in the EU to comply with GDPR.

(To meet these regulations, many turn to Compliance as a Service.)

Dissatisfaction with cloud services performance

One reason many companies turn to the cloud is because of its scalability. However, it’s not always the ideal environment for every type of workload. Some applications may experience lag issues when running on the cloud. This can be especially detrimental if they require real-time processing or have high data transfer requirements.

For example, remote surgery, real-time video streaming and self-driving vehicles all rely heavily on zero to minimal latency for optimal performance.

(Each of these issues can affect your CloudOps environment.)

The pros and cons of cloud repatriation

The cloud offers businesses many benefits. Likewise, transferring from the cloud comes with advantages and drawbacks. If you’re considering cloud repatriation, here are the pros and cons factors to consider.

Some advantages include:

Cost-effective

Managing workloads in the cloud can be more expensive than expected due to hidden costs, unpredictable pricing, and scaling difficulties. Repatriation helps reduce these unnecessary costs, particularly if the organization efficiently manages its infrastructure. One report found that cloud repatriation provides a 50% reduction in cloud spend.

Improved security and compliance

Cloud repatriation helps some organizations improve data security and meet compliance requirements. Plus, some leaders gain peace of mind and feel safer having sensitive data on-premises where they have direct control over security measures.

Better IT infrastructure control

Moving data back to an on-premises data center provides many organizations the complete control over the data and applications they want. This can be especially beneficial for security and compliance.

Reduced latency

Some applications perform better in an on-premises environment, particularly those that require real-time data processing with minimal latency.

To reduce latency, organizations might choose to take their data from the cloud and implement edge computing by building or leveraging edge data centers close to where data is generated.

However, some of the disadvantages of repatriation include:

Increased resource and management requirements

Many companies turn to the cloud to reduce infrastructure costs, so leaving the cloud will increase these costs. On-premises data centers require resources for their management and maintenance, including hiring or training staff to manage the infrastructure.

Lost scalability

One advantage of cloud computing is its elasticity, which allows businesses to scale their resources up and down as needed easily. Repatriating data to an on-premises data center may limit this scalability.

Initial transition difficulties and upfront costs

Transitioning back from the cloud can be challenging. Building or upgrading an on-premises data center requires a significant upfront investment. These costs include not only the hardware and software but also real estate, power, cooling, and more.

Is leaving the cloud right for you? It depends. Your decision should be based on thoroughly assessing your organization’s business goals, budget, security requirements, compliance needs, and technical capabilities.

Reassessing your cloud needs

While the cloud continues to grow, cloud repatriation represents an important shift in how organizations manage their data and application. As businesses strive to balance cost, performance, control, and security, moving away from the cloud back to on-premises or a hybrid environment can be an effective strategy.

Cloud repatriation is not an all-or-nothing issue. Instead, it’s a critical chance to take a more thoughtful approach to where you place your apps and workloads and decide if each one is serving you in the public cloud. Carefully consider the potential cost savings, improved control and compliance, and performance gains against the potential drawbacks, including increased resource requirements, reduced scalability, and higher upfront costs.

With increasing data generation and complex digital workloads, deciding where to store and process data has never been more crucial. Cloud repatriation is one of many IT tools that businesses can leverage to navigate this landscape successfully, ensuring their data management strategy aligns with their broader business goals and operational needs.

Related Articles

The Bulkhead and Sidecar Design Patterns for Microservices & Incident Resolution
Learn
3 Minute Read

The Bulkhead and Sidecar Design Patterns for Microservices & Incident Resolution

This article looks at Bulkhead and Sidecar design patterns, including how they’re used in microservice designs — and how they help overall incident support.
Content Delivery Networks (CDNs) vs. Load Balancers: What’s The Difference?
Learn
3 Minute Read

Content Delivery Networks (CDNs) vs. Load Balancers: What’s The Difference?

CDNs and load balancers fulfill similar roles, but they are different tools. This article breaks down the differences so you can decide which is right for you.
Best DevOps Books: The Definitive List
Learn
4 Minute Read

Best DevOps Books: The Definitive List

In this blog post we’ll look at the core, fundamental books that have played the largest role in creating the modern DevOps movement.
Kubernetes 101: How To Set Up “Vanilla” Kubernetes
Learn
4 Minute Read

Kubernetes 101: How To Set Up “Vanilla” Kubernetes

Kubernetes 101: Set up the most basic K8s cluster — also known as Vanilla Kubernetes — with this hands-on tutorial that gets you started quickly and easily.
Network vs. Application Performance Monitoring: What's The Difference?
Learn
5 Minute Read

Network vs. Application Performance Monitoring: What's The Difference?

Monitoring networks and application performance are different practices. Understand the changes and see how, together, both can offer end-to-end observability.
Monitoring Windows Infrastructure: Tools, Apps, Metrics & Best Practices
Learn
3 Minute Read

Monitoring Windows Infrastructure: Tools, Apps, Metrics & Best Practices

Learn how to monitor your Windows infrastructure, including the best tools and apps to use, the top metrics to monitor and how to analyze those metrics.
NoOps Explained: How Does NoOps Compare with DevOps?
Learn
5 Minute Read

NoOps Explained: How Does NoOps Compare with DevOps?

Take a look at NoOps, the concept of automating IT and development: how it works, pros and cons and whether it’s an evolution — or the end — of DevOps.
How To Prepare for a Site Reliability Engineer (SRE) Interview
Learn
4 Minute Read

How To Prepare for a Site Reliability Engineer (SRE) Interview

Prepare for your SRE interviews. These are common questions and answers to expect in any site reliability engineer interview.
Adaptive Thresholding with Splunk's Density Function
Learn
3 Minute Read

Adaptive Thresholding with Splunk's Density Function

Past data supports adaptive thresholding with Splunk. Learn how — and when — to use the probability density function to create adaptive thresholding.