What Is Cloud Migration? Importance, Benefits, and Strategy

Cloud migration is the movement of data, applications, computing capacity, and other digital assets from an organization’s on-premises data center to a data center run by a cloud provider. Like all migrations, the journey is driven by a search for greener pastures. An organization pursuing cloud migration seeks a place where computing costs are lower, the capabilities are numerous, new tech innovations arrive in regular releases, and scalability is almost unlimited. A move to the cloud seeks to end the pain of hiring talented technologists to perform repetitive work to keep hardware, databases, and applications up and running. But while benefits abound, the migration process—if poorly planned—can be arduous, even perilous. The information here will help you understand what’s at stake with cloud migration planning and help you make the right preparations to speed you on your way to this new and better place.

What Is Cloud Migration?

Cloud migration is the process of moving applications and their supporting technologies out of an organization’s on-premises data center and into a cloud provider’s data center. This can mean moving an on-premises application along with supporting databases and development platforms onto the cloud provider’s servers, storage, and networking infrastructure. It can also mean moving to a cloud application, such as an ERP, CRM, or human capital management (HCM) application delivered as software as a service, and shutting down the legacy version.

In all these models, cloud migrations shift workloads out of an organization’s owned data centers and into a cloud provider’s data center where tasks such as backups, security patching, and new feature releases are generally a regular part of the service. Using cloud services is usually less expensive compared with running a data center because you only pay for what you use. As migrations have become more common, companies have discovered another benefit: Migrating frees skilled employees from the mundane work of “keeping the lights on” in the data center and allows them to focus on supporting new business opportunities.

The term “migration” is used because moving these workloads can be a significant undertaking that requires detailed planning and deft execution. Such planning is essential to keep operations running while an organization cuts over to cloud-based systems.

Key Takeaways

• Using cloud resources is an increasingly common business strategy.
• Cloud migration strategies rely on a wide array of proven processes and tactics.
• Successful migrations require caution and planning, but the risks are well understood.
• There are a growing number of tools from cloud providers and third-party vendors that make the process manageable.
• Cloud providers continue to innovate and improve their services, making the cloud an enticing option for even critical and sensitive workloads.

Cloud Migration Explained

Cloud migration involves taking existing databases, applications, and other computing workloads that are running in an on-premises infrastructure and moving them to a cloud computing provider. This could involve moving hundreds or even thousands of databases; applications for critical functions such as finance, billing, human resources, supply chain, or manufacturing; or development environments used by teams who create and test new apps.

In the early days of the cloud, migrations weren’t as common as they are now. For more than a decade, running IT operations in the public cloud was a strategy mostly applied to new applications or digital assets. Using a cloud environment, a startup could bring an idea to market quickly without having to install and maintain servers and buy software licenses. They could simply pay for the cloud infrastructure they used and get going right away. Since the public cloud means sharing computing resources with other companies, however, more-established companies and governments initially worried about privacy and data security issues, and about other customers who might reduce application performance by hogging CPU cycles. As a result, more-established companies and governments with their own data centers, especially those running critical applications, tended to shy away from public cloud offerings.

Over time, the cloud landscape has changed dramatically, and cloud migrations have become commonplace. Cloud providers have upgraded their systems and delivery models, and they’re offering infrastructure options that appeal even to those with the most stringent privacy, security, performance, and reliability needs. Now organizations in fields such as global finance, healthcare, communications infrastructure, and government see the cloud as a way to get the latest innovations by using cloud services that their own data centers and software versions can’t offer. Cloud costs are generally lower than what it costs companies to run their own infrastructure. The question then becomes how to move data and applications from their own data centers to the powerful, fault-tolerant, and up-to-date data centers of a cloud provider.

How Does Cloud Migration Work?

Cloud migration is the process of transferring data, services, and applications to a cloud provider’s data center. It typically involves a detailed plan for backing up data, understanding network connections, and ensuring data security during the migration process. It also involves working with the cloud provider to understand what application changes, if any, are needed for the application to work in the new cloud environment. Migration requires mapping an architecture that fits with the cloud provider’s infrastructure, including the cloud provider’s methods for maintaining data privacy and security, such as their authentication methods to protect against cyberattacks.

To ensure a successful cloud migration, organizations often enlist experienced cloud architects and technicians who are familiar with the cloud provider’s platform to ensure consistent backup, data portability, and application compatibility between the cloud and the on-premises data center the organization is leaving. These specialists can also explain how cloud-based hardware and software can accelerate data processing. If done properly, and in consultation with industry experts where needed, migrating data and applications to the cloud can be done with minimal disruption.

Migration Deployment Models

One of the key reasons why cloud migration is becoming more common is that cloud computing offerings have become more diverse and sophisticated. The largest cloud providers—sometimes referred to as hyperscale providers—now offer a wide array of options that go well beyond public cloud services, including the following four deployment models:

• Public cloud is the traditional model. Organizations pay for space on servers in the cloud provider’s data center, either based on usage or by subscription. The servers are shared by that cloud provider’s customers.
• Private cloud is where the whole server or specific server racks at the cloud data center are dedicated to one organization. Depending on the customer’s needs, some cloud providers may put the dedicated hardware in the organization’s own data center while retaining responsibility for its management.
• Hybrid cloud means an organization moves some data and applications into a cloud provider’s data center and keeps some in their own data center, creating a shared “hybrid” architecture that supports the applications. An organization might choose this model because of company policies or data residency regulations.
• Multicloud is the process of migrating applications and data to different cloud providers. This allows an organization to run workloads with the cloud provider that specializes in that kind of service. For example, an organization might run its application server on one cloud and database operations on another.

Migration Service Models

There are a few different ways to approach cloud migration based on what an organization is trying to achieve. They range from migrating to a cloud-hosted enterprise application, such as an ERP or CRM application, to leasing core computing infrastructure and migrating a current application or your development-and-testing environment onto it. Here are the three main models.

• Software as a service (SaaS)
  SaaS cloud migration means taking the data and functionality of an application running in your data center and moving the entire business process so it runs on a cloud-based application from the cloud provider. For example, you might currently run accounting processes using an on-premises general ledger accounting application that’s deployed and maintained by your IT department; SaaS migration would mean moving to an accounting application that’s written and continuously patched and updated by the cloud provider, that runs in the provider’s data center, and that your employees access using a browser.
• Infrastructure as a service (IaaS)
  IaaS migration is often referred to as “lift-and-shift” migration. This migration model takes an application from your own data center and moves it—largely as is, ideally—onto compute, storage, and networking infrastructure in a cloud provider’s data center. The cloud provider handles all maintenance, updates, and patches for that hardware.
• Platform as a service (PaaS)
  PaaS cloud migration means using cloud services—such as databases, application servers, operating systems, containers, and development tools—that let you build and run new applications in the cloud provider’s data center. This includes core infrastructure, such as servers and networks, and also platform services, such as database, integration, and possibly analytics services. This gives you control and the ability to build specific functionality for your application on a scalable architecture.

Types of Cloud Migration

Cloud migration is a single term that covers a wide array of business-technology strategies. Here is a list of six cloud migration options and why you might choose them.

1. Rehosting sometimes goes by the name “lift and shift.” In this process, you move an application to a cloud provider’s data center with as few alterations as possible. A lift-and-shift application continues to deliver the same capabilities to your employees or customers; ideally they never even know it moved. The difference is instead of being hosted in an on-premises data center it’s now hosted in a cloud data center, so it’s running atop the latest hardware technology, which is maintained by the provider, not your IT staff.

2. Replatforming also takes an on-premises application and moves it to a cloud provider’s infrastructure. In the replatforming process, however, the application is updated to take advantage of new technology or services available from the provider. For example, when moving an HCM application from your own data center infrastructure to the cloud, you might replace a labor-intensive and older data management environment with an autonomous database that does automatic updates and offers built-in machine learning models.

3. Repurchasing replaces an application running in your on-premises data center with a cloud-based, provider-created SaaS application that employees access via a browser. For example, you might move from a licensed, on-premises ERP application to an enterprise ERP cloud service that you pay for by subscription and that is automatically updated several times a year with new features. While this sounds simple, it takes time and planning to map the new application’s functionality to the processes you use to do business. It may require organizational change to get your team to adopt what are considered industry best practices built into the cloud applications. Cloud providers offer tools to help you map the conversion.

4. Refactoring is the process of moving an application to the cloud with the intention of modernizing its architecture to take advantage of cloud native features. For example, a monolithic application built over time by your organization might still do the job it was built to do, but it may be too hard to add new features to address evolving customer needs or go after new opportunities. By refactoring the application, you can introduce a microservices architecture that makes it much easier to develop, test, and launch new features to such an app. Or you might add in-database analytics that make it easier to run analytics without moving data around your environment and help you get more value from your data.

5. Retaining means you’ve looked closely at what it will take to move an application to the cloud and have determined that, for now, a move doesn’t make sense. There can be a lot of reasons for this. Maybe your application has low latency requirements that favor an on-premises model, or you need to follow data residency rules that make you wary of moving to a cloud provider’s data center. Or perhaps, after doing your homework, you realize the cost and effort of the migration outweigh the benefits—at least for now. Whatever your reason for deciding not to move to the cloud, it's still wise to revisit the idea from time to time. Cloud providers continue to build data centers in regions around the world, add new models that address data control, and improve the efficiency of the migration process.

6. Retiring happens when you look closely at an on-premises application and realize its functionality is hardly used or no longer needed. Saying goodbye to these apps can deliver a big win from the cloud migration process because you’re eliminating redundancy or vestigial processes that cost you money but no longer deliver value. Retiring an application does take time and planning because there can be dependencies with other applications that need to be addressed before you turn it off.

The Benefits of Cloud Migration

The reason cloud migration is an increasingly popular business strategy is because cloud-based applications, platforms, and infrastructure can offer more benefits at a lower cost. Here are some of the benefits of migrating to the cloud.

• A tough critique of your own tech environment
  The process of cloud migration begins with a thorough review of your IT architecture, applications, and data management, which is helpful regardless of whether you end up moving a particular asset to the cloud. Asking “Is it worth moving this app?” can force you to make the tough call to retire an app that’s past its prime.
• AppDev agility
  Cloud-based development environments let teams launch a DevTest setup and get coding in minutes instead of the days or weeks it could take if developers need to order and provision new hardware.
• Elastic scalability
  Cloud-based infrastructure can be scaled up or down quickly to meet fluctuating business needs. This negates the need to overprovision hardware to meet those very rare peaks in demand, and it lets organizations slim down their cloud costs quickly when resources aren’t in use.
• Pay-per-use model
  Cloud-based infrastructure is purchased by time used or CPU usage. This saves money compared with on-premises architectures for which hardware and software licenses must be purchased and provisioned up front to accommodate the organization’s maximum needs, even if that level of need is rarely reached.
• Self-service provisioning
  Cloud services are often designed to be provisioned by end users without involving in-house technical expertise. For example, a developer who needs a test database can provision an autonomous cloud database in minutes instead of having to wait for a database administrator to find the time—and hardware capacity—to provision it.
• Constant innovation
  Cloud providers keep their data centers on modern hardware that’s patched against the latest security risks; they also update their cloud-based applications and infrastructure automatically and offer regular updates with the latest features. Meanwhile, on-premises applications often run for years without a significant upgrade.
• More value from data
  Cloud platform services offer the most up-to-date versions of databases, which can offer features such as the ability to manage multiple data types and deliver machine learning in the database engine—vastly simplifying the architecture needed to get value from data.
• Lower data center costs As businesses adopt cloud-based resources and move data, applications, and development platforms to the cloud, they can close or largely decommission their expensive data centers. Another benefit to closing data centers is that it takes the burden of routine tasks off skilled employees, freeing them up to do work that adds new business value.

The Challenges of Cloud Migration

Even with the long list of potential benefits, some applications might not be candidates for the rigors or risk of a cloud migration. Or, at least, it will take extra research and planning to move them. For example, it may be more challenging or riskier to move business-critical applications, high-throughput applications that require low latency, or applications that have strict data residency requirements. The risks of cloud migration are fairly well understood, however, and don’t have to be deal breakers for most workloads. Let’s look more closely at the challenges of cloud migration.

• Network latency
  For some applications with very high throughput, such as financial trading or telecommunications applications, the time it takes to send data over the internet to a cloud provider, a period called latency, can be an issue that needs to be planned for and addressed. But for most applications, the impact isn’t noticeable.
• Data residency
  Some data, whether due to company policy or data residency regulations, can’t be moved to a cloud provider’s data center. However, these objections can sometimes be addressed if the cloud provider has a data center inside the company’s country or region, or by putting cloud infrastructure that the provider manages inside the company’s data center.
• Complexity
  For companies with hundreds or thousands of databases and applications, moving data and business processes to the cloud can be a complex undertaking that requires thorough discovery and planning and deft execution. Often such large-scale migrations are done in phases.
• Data and application security
  Companies spend much time and effort building security processes around their on-premises applications and databases. When migrating to the cloud, they must complete the necessary audits and assessments to know that their data is secure during the migration and once it’s in the cloud provider’s data center. This process can be more complex when migrating to a hybrid cloud configuration where data regularly moves between the customer’s data center and the cloud provider’s data center. But the reality is that cloud providers can invest much more in security expertise, monitoring, and maintenance than companies can themselves.
• Organizational changes
  When migrating to a SaaS application, the application will have assumptions built in about the business practices it supports. These are often based on assumed industry best practices. An organization might have to alter a business practice to take advantage of the new application, which can require a change management process.

Six Cloud Migration Steps

Companies migrate to the cloud for different reasons and at different scales—from small testing platforms to large-scale migrations. The basic process, however, requires the same steps, which include identifying what can be improved by moving to the cloud, taking inventory of the supporting data and infrastructure, and considering if or how they can be replicated or even re-architected using cloud services. The goal of your cloud migration program is to get the agility and cost savings of the cloud while keeping data and application processes secure, both during and after the migration.

Here are six steps your cloud migration process should include, using a SaaS application migration as an example.

1. Discover
Use software monitoring and management tools that help you get a clear picture of your data and application infrastructure and its dependencies and policies. You’ll need to replicate or re-architect these in the cloud.

2. Design
Designate which cloud services will support or replicate functionality and the process you’ll use to securely migrate to them.

3. Build
Use cloud resources and services to build out the new application and infrastructure architecture, including the pathways you’ll use to migrate data securely.

4. Integrate
Replicate the data in the new application, and make sure it integrates with all existing data sources and analytics processes.

5. Validate
Run and test the new application to confirm it functions as desired.

6. Cut over
It’s the day you’ve been working toward. Bring the new application into production and invite some users to test it. Eventually you can bring all users over to the new cloud-based application.

Cloud Migration Tools

Cloud migration can be a complex undertaking, but there are a growing number of tools that are designed to make it more manageable. Cloud providers, such as Amazon, Google, Microsoft, and Oracle, often provide these tools. Migration tools are also available from third-party vendors, such as AppDynamics, Carbonite, and CloudScape, that specialize in the discovery, networking, and reporting processes required. Each tool is designed to help with one part of the process of migrating data, applications, or infrastructure from on-premises to the cloud—with the least disruption to business operations or application downtime. Here are some of the tools that address common migration challenges.

• Discovery and translation tools look deep into an application’s architecture to identify data sources and hardware and software dependencies. These tools can then use that information to help the organization understand which cloud services are needed to replicate a process.
• Network and resource discovery tools seek to give an organization a clear understanding of how their team members use an application and its services. The tools can then help map that required functionality to an application or database in the provider’s cloud.
• Migration and installation tools provide a means to shift data, applications, network configurations, and other services to the cloud provider’s servers and validate that they’re working properly.
• Reporting and graphing tools help stakeholders track and validate the migration during the migration and after the cutover.

Cloud Migration Tips and Best Practices

Whether you’re migrating a database, moving to a SaaS application, or adopting cloud-based compute and storage services, these steps and best practices can guide you.

1. Clearly define the scope of your project
For example, know exactly which applications or databases you will migrate and why.

2. Discover dependencies and integrations
There are cloud migration tools that can help you map the data sources, security processes, and dependencies. Only after locking down this information can you recreate them in the cloud.

3. Find ways to add value
Thoroughly review the cloud services offered and look for opportunities to take money-saving steps, such as automating database operations or simplifying analytics processes.

4. Map your business processes to the new application
Before you migrate, understand your current business processes related to the technology you plan to move. Next, work to understand if and how the cloud destination will force you to change your work processes. For example, if you’re moving to a SaaS application, it might assume best-practice work processes that are different than what you used with your legacy application.

5. Back up everything
Need we say more?

6. Focus on data security during and after the migration
Understand how your cloud provider approaches security, and be ready to take advantage of their expertise and capabilities. Understand what security-related configurations and updates are your responsibility and what the cloud provider handles.

Cloud Migration Trends

The most significant recent trend has been organizations choosing to move increasingly demanding workloads to the cloud and to pursue more-complex and sophisticated hybrid cloud and multicloud migrations to accommodate those workloads. Expect some cloud providers to fight against the multicloud trend by imposing high data egress fees and other obstacles, and others to embrace it by working with fellow cloud providers to simplify and lower the barriers to multicloud deployments.

Another trend is the growing interest in platform services, such as cloud native development platforms, which allow companies to build their own software while getting the benefits of state-of-the-art, pay-as-you-go cloud economics.

Finally, a hot trend to watch is the growth of cloud deployments in a customer’s own data center. These dedicated clouds reside in a customer’s data center but use hardware and software that’s owned and operated by the cloud provider. The customer pays only for the cloud services they use, just as they would if they were using resources in the cloud provider’s data center.

As with most technology trends, look for cloud migration to grow more sophisticated as cloud providers build even more enticing offerings and continue to simplify the migration processes.

Make the Move to the Cloud with Oracle

Cloud migration is no longer a cutting-edge strategy—many organizations, both huge and tiny, have successfully moved to the cloud. It does take planning and effective management, but the risks are well known and the process for a successful migration is well established. Companies that haven’t made the move should consider moving a small part of their operations to start, but the biggest gains will come from moving major operations to the cloud so they can close down their data centers altogether. Cloud migration promises lower costs, self-service IT resources, elastic scalability, automation, and constant innovation. And although the process can be complex, automation tools and cloud provider resources make it a simpler and more secure process. These forces combine to make cloud migration a significant opportunity for IT organizations.

When it’s time for your organization’s migration to the cloud, Oracle Cloud provides a compelling suite of cloud applications, infrastructure, and platform services. Oracle Cloud Infrastructure was designed and built from the outset to handle the most-demanding workloads and to make it easy to move data and databases to the cloud. Whether your organization’s looking to adopt a suite of cloud-based enterprise applications, move your existing applications to a public cloud or multicloud infrastructure, or run cloud workloads in your own data center, Oracle Cloud has services to meet your needs.

Cloud Migration FAQs

What are the types of cloud migration?

Cloud migration types include rehosting, sometimes called lift and shift; repurchasing, which replaces an application running in your on-premises data center with a cloud-based application accessed via a browser; and refactoring, which is the process of moving an application to the cloud with the intention of modernizing its architecture to take advantage of cloud native features.

What are the four main phases of a cloud migration?

There are many steps involved in a cloud migration, but the four main ones are (1) discover, where you develop a clear understanding of the workload being migrated; (2) build, where you mirror existing functionality in the cloud data center; (3) validate, where you test that all the pieces work as planned; and (4) cut over, where you bring users onto the new, cloud-based services.

Why migrate to the cloud?

Organizations migrate IT workloads to the cloud to get benefits that include lower costs, greater agility, elastic scalability, more automation, self-service provisioning, and constant innovation.

What are the top challenges of cloud migration?

A key challenge is the high complexity of moving workloads from an on-premises data center to a cloud provider’s data center, in particular understanding all the dependencies and addressing data security concerns. Network latency issues for high-throughput operations and data residency concerns can be additional obstacles.