The rise of cloud repatriation
Explore cloud repatriation where businesses exit the public cloud and embrace on-premises, private cloud, and hybrid solutions to reduce expenses.
The BPM market is projected to reach $65.8 billion by 2032. A key reason for the growth is the increased adoption of cloud-hosted solutions with migration from on-premises often associated with cost savings.
However, Gartner have revealed that companies are overpaying by up to 70% for their cloud infrastructure.
In this blog, we examine the growing trend for businesses to exit the public cloud and embrace on-premises, private cloud, and hybrid solutions to reduce cloud expenses, meet compliance mandates, and improve data security.
What is cloud repatriation?
Cloud repatriation (also referred to as “cloud exit” and “reverse cloud migration”) is when an organisation moves applications and data from the public cloud (e.g., AWS, Azure, Google, and IBM) to a private cloud, hybrid cloud, or on-premises setup.
What is the public cloud?
Public cloud refers to computing services (e.g., servers, storage, and applications) provided to customers over the Internet.
Public cloud infrastructure is shared between multiple customers (tenants) and is available as a pay-as-you-go model, meaning you only pay for the resources you consume. This enables organisations to mitigate the risk of overprovisioning or missing capacity and quickly adapt to changing market conditions.
The main providers of public cloud services include Amazon Web Services (AWS) and Microsoft Azure (also referred to as hyperscalers).
What is the private cloud?
A private cloud is where computing resources/infrastructure is dedicated to one client.
This can be on-premise, or within your own or shared data centre. With shared data centres, the provider usually only supplies rack space, power and a connection to the Internet with the owner responsible for the management of all software and hardware.
There are many providers of private cloud services in the UK including Redcentric, Datum or Equinix, for example.
Why consider cloud repatriation?
According to Techopedia, over 43% of IT leaders found that moving applications and data from on-premises to the cloud was more expensive than expected.
At the same time, Citrix reports that 42% of organisations surveyed in the USA are considering or have moved at least half of their cloud workloads back to on-premises infrastructures.
Some of the reasons why businesses are repatriating workloads from public to private clouds (or on-premises) include:
Complex pricing and billing
The pricing strategy of public cloud providers is complex and difficult to understand with costs varying depending on the location and pricing tier. To illustrate, one of Amazon’s most popular cloud products is EC2 and there are multiple ways to pay - on-demand, savings plans, reserved instances, and spot instances. EC2 is also available on the AWS Free Tier with 750 hours per month for 12 months.
Escalating costs are another reason for Cloud repatriation. As your capacity needs increase, costs can spiral out of control, especially if resource usage is not carefully monitored and optimised.
Additionally, public cloud providers promise cost savings, but many organisations find the savings never materialise. For example, David Heinemeier Hansson, the co-owner and CTO of 37signals, has written a blog post on LinkedIn documenting how the makers of the Basecamp project management tool have shifted their seven cloud apps from AWS into their hardware at a professional data centre. In the article, David says their cloud bill is now $1.3 million compared to $3.2 million with AWS, a saving of $1.9 million.
The hyperscalers have also been accused of introducing price increases when launching new features with customers surprised by an unexpected bill.
Many users complain about the difficulty of interpreting their bills and understanding the resources used in the billing period.
Vendor lock
An Ofcom review into the UK cloud market has found it is difficult for UK businesses to switch and use multiple cloud suppliers with AWS and Azure having a combined market share of 70% to 80% in the UK in 2022.
One of the main concerns is data egress fees. Organisations pay egress charges to transfer data away from cloud providers. These fees can discourage customers from switching to private cloud or on-premises, leading to vendor lock and preventing organisations from adopting a “multi-cloud” strategy. Technical barriers including limited interoperability and portability also contribute to vendor lock.
The pricing strategy of the hyperscalers encourages vendor lock with savings for long-term commitment and volume discounts for increased usage structured to incentivise customers to use a single provider.
These concerns have led Ofcom to refer the public cloud infrastructure services market to the Competition and Markets Authority for further investigation.
Product ecosystem/service sprawl
Hyperscalers have hundreds of products and solutions each with their own features, configurations, and use cases. This is a challenge for customers who must navigate an ever-increasing product portfolio. With several overlapping products, customers can get confused when choosing the best solution for their needs.
Designing cloud computing services like Amazon EC2 or Azure Virtual Machines requires an in-depth understanding of cloud provisioning with misconfiguration leading to performance issues, higher bills, and security breaches.
For instance, hackers exploited misconfigured AWS systems to compromise over two terabytes of data including customer information, infrastructure credentials, and proprietary source code.
Regulatory compliance
Highly regulated organisations, such as those in the financial services and the pharmaceutical industry, face specific challenges. They are subject to more scrutiny than their lightly regulated counterparts.
For instance, when using the public cloud to host sensitive data, information must be tightly protected with strict guidelines on who has access to it, where it is stored, and how it is processed.
Consequently, repatriation to on-premises or private cloud is a good option for organisations in regulated industries that seek greater control.
Data security
Public cloud infrastructure is shared between multiple customers (tenants), which can introduce potential vulnerabilities.
Moving data away from the public cloud to an on-premises solution allows for greater control over data privacy, security and compliance to local legislation (e.g., GDPR in the EU).
Performance issues and concerns
Public cloud can be less than ideal for organisations that require high performance computing with minimal delays (known as low latency). Migrating to an on-premises solution with dedicated hardware can improve performance and reduce latency issues.
Constant workloads
Applications running 24/7 with stable utilisation patterns are more expensive in the public cloud than on dedicated resources on-premises. Organisations can realise significant cost savings by repatriating them.
Successful cloud repatriation
Moving workloads from the public cloud back to on-premises or private cloud can be challenging and requires careful planning. Our recommendations include:
Conduct an in-depth assessment
Identify suitable workloads for cloud repatriation and quantify the potential cost savings. You will also need to plan the physical infrastructure (e.g., rack space, power, storage, and cooling) and create a team to support the migration and day-to-day management.
Develop a migration plan
Create a detailed roadmap with steps for data migration and application transition, including timelines, resource allocation, and contingency plans. This will allow for a smooth transition.
Consider hybrid cloud
Consider hybrid cloud environments that combine on-premises with public and private cloud infrastructure. For example, on-premises is better for mission-critical applications or organisations that must adhere to legislation. Conversely, the public cloud is ideal for unpredictable workloads that may require instant up-scaling.
Seek expert help
Third-party consultants and managed service providers can provide invaluable support before, during, and after repatriation. Their skills and knowledge can reduce risk - accelerating the migration of data or applications and providing support and training to your team post-repatriation.
Managing public cloud costs
For organisations committed to the public cloud. Here are some best practices to manage costs:
Build incrementally
We recommend starting small with the public cloud, for example, by provisioning a core workload on an Azure VM. The result is a quick win, allowing you to get up to speed with creating and managing virtual machines in Azure without the risk.
As you become more familiar with the environment, you can extend your cloud capabilities and deploy additional Azure products such as SQL Databases and Azure Storage, to suit your requirements.
Monitor costs
Cloud costs are based on usage and could increase exponentially if your demand increases over time. This means you need to monitor costs to reduce the risk of “bill shock”.
Public cloud providers provide cost management tools to enable customers to manage cloud costs better. These tools allow organisations to analyse usage, monitor spending patterns, and optimise costs – but their existence suggests the hyperscalers know a problem exists!
For instance, you can set up budget alerts in Microsoft Azure to provide advanced notification if spending is likely to exceed your budget. These tools are a good way to start monitoring expenditures but are limited to the cloud provider - and their objective is to increase profits.
Sometimes these cost management tools can be lacking, leaving the customer to invest significantly in trying to monitor, understand and predict their usage charges. As an example of this, Netflix, a well-documented early adopter of the AWS public cloud, is still striving for “cost insight” over their cloud infrastructure.
There is a market for third-party extensions that integrate with the hyperscalers cost optimisation tools and provide extra functionality to keep costs under control. Similarly, there are hundreds of companies providing consultancy services to cloud customers promising to reduce cloud bills without sacrificing resources or performance.
We recommended reviewing contracts and paying attention to the costs (egress fees) of extricating the organisation from the contract.
Take advantage of discounts…but tread carefully!
AWS, Azure, and Google provide multiple pricing options for customers including on-demand, savings plans, reserved instances, and spot instances.
These pricing models allow customers to benefit from volume-based discounts and save when usage increases. Commitment-based pricing offers savings for using a specific amount over a fixed period (usually one year or three years).
However, using discounts does have significant disadvantages. You are committed to the price even if the public cloud provider decides to cut prices and commitment-based pricing encourages vendor lock.
Beware of free gifts
Many providers have generous initial offers, providing free credits and discounts while you evaluate and transition to their public cloud. However, be aware that whilst this allows for a risk-free trial, it also obfuscates the ongoing costs of the service and makes it hard to understand what you will be paying once these introductory offers expire.
Reduce usage with right sizing
Inefficient allocation and underused resources such as leaving an old server running will increase costs needlessly.
Right sizing is an effective way to control cloud costs by provisioning instances to match your actual workloads. It involves continually monitoring performance, usage requirements, and trends over time with IT professionals removing unused instances and adapting instances that are overprovisioned or risk missing capacity.
You can automatically monitor utilisation patterns with cost management tools, enabling you to reduce costs without compromising workload performance.
Consider a multi-cloud strategy
A multi-cloud approach is using cloud computing services from two or more cloud providers to run your applications. According to OVHcloud, 64% of organisations see their use of multi-cloud increasing in the next two years.
Adopting a multi-cloud approach can offer greater flexibility to customers and avoid vendor lock. Multi-cloud environments can also reduce IT expenditure with cloud providers offering different pricing models, enabling organisations to select the most cost-efficient vendor for specific workloads.
But every silver lining has a dark cloud…
However, not everyone is convinced by a multi-cloud with some arguing single-cloud strategy offers simplicity. Netflix and Spotify seem to agree, running exclusively on AWS and Google Cloud respectively.
Each cloud provider has a bewildering array of services and offerings, which can become a significant learning curve to fully understand and operate. Moving to a multi-cloud approach not only complicates your IT infrastructure, but also your staff recruitment, training, and the skillsets required.
An increase in providers also increases the surface area of your systems to the Internet, which increases the security risks and work required to ensure it is safe.
Finally, as mentioned above, cloud vendors often provide volume-based discounts, so by spreading your workload across more than one vendor you minimise the discounts available to you, leading to significantly higher costs.
Conclusion
The benefits of the public cloud, such as its scalability and elasticity, are well understood. Yet, the drawbacks – chief among them controlling costs and ensuring regulatory compliance – are driving the gradual trend of cloud repatriation.
There are pros and cons to both private and hybrid clouds.
With private cloud migration, you do not need your own data centre. You can work with a managed service provider and rent racks, bandwidth, and power which will cost far less than spend on public cloud providers like AWS, Azure, and Google. Ultimately, third-party data centres bridge the gap between expensive public cloud services and on-premise or in-house data centres, providing a dedicated cloud platform with the connectivity and security you need at a competitive price.
The hybrid cloud, meanwhile, allows organisations to create a flexible infrastructure with sensitive applications managed on-premise (or private cloud) and workloads suited to it, running on the public cloud.
Ultimately, we recommend deploying workloads in the most appropriate location for your organisation depending on cost, resource utilisation, data security, and compliance requirements. Hopefully this blog will help you fully understand the factors to weigh up before making that decision.
To discuss your cloud repatriation strategy, please call us on 03300 100 000 or complete this form.