What is a Cloud Migration Strategy?

A cloud migration strategy provides a business case, scope, and strategy for migration from current IT resources to a new cloud environment. This strategy outlines each existing component, such as virtual private network infrastructure, an internal application, or a development environment, and the best method to migrate it to the cloud for a balanced business benefit.

Cloud migration benefits can include increased performance and scalability, reduced operational costs, and enhanced operational efficiency and resiliency. Cloud adoption expedites business opportunities and technical outcomes while reducing business risks through improved governance and security.

However, migrating to the cloud isn’t a simple transfer process. Common cloud migration challenges include:

• Securing executive support and other stakeholder buy-in
• Discrepancies between infrastructure types
• Cloud provider configuration difficulties
• Learning new systems
• Security and cost management considerations in transfer and configuration
• Determining hybrid needs between on-prem and cloud resources

A successful cloud migration strategy helps maintain continuous business operations. Beyond that, planning is necessary to implement a highly efficient cloud architecture at low costs for future scaling and integrability.

Which cloud models are considered in a cloud migration strategy?

Organizations can choose a number of different cloud models for their cloud migration strategy:

• Public cloud: Where you consume services over the internet from a cloud service provider (CSP) running their own infrastructure
• Private cloud: Where you run services from a CSP on-premises
• Hybrid cloud: Where you run a combination of on-premises and public cloud services
• Multi cloud: Where you run services in multiple public cloud from various CSPs

In practice, many organizations now run a hybrid or multi-cloud environment to meet their business objectives.

What are the cloud service delivery models?

CSPs offer three core cloud service delivery models, to suit various IT needs:

• Software as a Service (SaaS): A software application in the cloud, for example Amazon Quick Sight, a business analytics application
• Infrastructure as a Service (IaaS): An IT infrastructure component in the cloud, for example Amazon Elastic Block Store, for block storage of digital assets
• Platform as a Service (PaaS): An environment and runtime for development in the cloud, for example AWS Elastic Beanstalk, a full-stack application development cloud platform

A successful cloud migration strategy helps get you up and running with as little disruption to business operations as possible and with minimal unexpected challenges. To start building your strategy, first audit the cloud capabilities and configurations of the cloud infrastructure you choose. Also, factor in:

Security and compliance requirements

Conduct a security assessment before your migration to identify any potential risks to your sensitive data. You must also understand and outline how security will operate within your cloud environment, including identity management and access controls.

Your new cloud environment comes with different governance considerations than on-premises data center infrastructure. Establishing a new governance framework before you migrate anything will help your team enforce all policies and compliance with regulatory requirements, during and after the transition. For example, your on-premises infrastructure is located alongside your business residence, whereas you can choose the location of each of your cloud resources. Location in particular has implications for compliance obligations such as HIPAA and PCI-DSS.

Automation and migration tooling

Automation reduces manual effort and also supports infrastructure-as-code (IaC) development, helping make sure that configurations remain consistent across environments. Organizations should assess the proper tooling needed to optimize their migration speed and reliability. AWS Transform is a good place to start building a cloud migration roadmap.

Training for on-premises to cloud migration and post-migration operations

A successful cloud transition depends on the readiness of both IT teams and end-users. Your training programs should cover the functions of the new cloud technologies, security protocols, and operational best practices. Establish a knowledge base for documentation and troubleshooting, and make sure that you have adequate ongoing support resources.

Scheduling and timelines

A migration timeline for a phased migration helps minimize disruptions. Your timeline can prioritize critical workloads first to help maintain operational continuity.

Beyond the transfer of resources, your cloud migration plan should include a schedule for decommissioning legacy infrastructure, redundant licenses, and securely archiving or deleting old data.

Incremental transformation and ongoing optimizations

An incremental transformation strategy is a common choice that many organizations use during their cloud adoption. Gradually migrating applications introduces more checkpoints for engineers to validate their new cloud resources. After migration, regular monitoring helps make sure that cloud resources are used efficiently.

Optimizing costs and cloud spending

A central component of a successful cloud migration strategy is transitioning your accounting model for infrastructure from capital expenditures (CapEx) to operational expenditures (OpEx). By adopting a pay-as-you-go pricing model, businesses can remove the need for large upfront investments in on-premises infrastructure and instead pay only for the cloud resources they consume. In many instances, OpEx cost expenditure is a more effective resource allocation strategy.

Effective cost management of your cloud spending will involve regular monitoring to identify potential waste. In cloud resources waste typically takes the form of underutilized instances or redundant data storage. Using techniques such as rightsizing, auto-scaling, and reserved instances enables teams to minimize migration costs and maximize the ongoing business value of their cloud investment.

AI and ML cloud computing environment integration

AI and ML integration and analytics can be a major business driver behind cloud migration. This means first understanding the best cloud architecture and services for direct integration with AI and ML services, to deliver new applications to your business. The best-fit cloud architecture and services for AI and ML adoption can help determine your cloud migration strategy for specific resources.

Different resources need different approaches to migration. There are seven common migration strategies for moving existing applications and services to a cloud computing environment, known as the 7 Rs: Retire, Retain, Rehost, Relocate, Repurchase, Replatform, and Refactor (Rearchitect).

Selecting the appropriate migration strategy depends on the specific requirements of each resource, your IT environment, and the desired business value you want to generate.

1. Retire - When a barely-used resource is archived

You will choose a retirement strategy when a resource is no longer needed. It involves decommissioning or archiving application servers that no longer provide business value or use unsupported legacy technology.

Perform an IT landscape assessment

You can perform an IT landscape assessment to help identify workloads with extremely low resource utilization, which often indicates a server or other resource you can retire. For example, applications with very low CPU and memory consumption are typically prime candidates for retirement. Workloads with low consumption that have received no inbound requests over several months are also frequently considered for retirement, so they don’t incur migration costs unnecessarily

2. Retain - When a resource remains on-premises

Retaining a resource means keeping it in its existing environment without migrating. Organizations will often choose to retain existing infrastructure that they recently upgraded with significant capital investment. They may also choose to retain third-party legacy applications held on-premises until the vendor releases a compatible SaaS version. Regulatory requirements or complex physical dependencies may also necessitate retaining a resource on-premises.

3. Rehost (Lift and shift migration) - When a resource moves directly to the cloud

Often called lift and shift migration, rehosting involves moving a physical or virtual server’s configuration and workloads to a cloud environment with minimal changes.

This is a common on-premises to cloud migration for organizations that want a fast transition to the AWS cloud without performance disruptions. It works best for stable existing applications where the goal is to reduce capital expenses quickly without changing the core architecture.

4. Relocate - When a resource moves to an equivalent cloud structure

The relocate strategy moves infrastructure from one environment to another when the target cloud infrastructure offers equivalent mappings for existing resources.

An example is relocating multiple virtual servers running on an internal VPN to a virtual private cloud (VPC) structure. You can also use this approach to reconfigure cloud-based systems by moving instances or objects between different internal AWS accounts, regions, or Virtual Private Clouds (VPCs) to maintain business operations. For example, you transfer an Amazon RDS DB instance to another AWS account.

5. Repurchase - When you change a resource’s licensing to fit a cloud model

Repurchase, historically referred to as “drop and shop,” involves moving from your current application to a different cloud native service, typically a SaaS platform.

Common use cases include replacing an on-premise license with a SaaS license or replacing a custom-built legacy system with an out-of-the-box solution from a cloud service provider. Organizations typically assess these new products against specific compliance requirements and security measures before the transition.

6. Replatform - When you change your resource slightly to take advantage of cloud efficiencies, costs, or capabilities

Replatforming, commonly referred to as “lift, tinker, and shift,” or what AWS officially terms “lift and reshape,” means performing small optimizations to an application that allow it to take advantage of new cloud capabilities without changing its core architecture. You see this approach used by legacy applications teams that want to scale their applications without risking the introduction of new security concerns.

Examples include moving virtual machines to containers, migrating a Microsoft SQL Server database to Amazon RDS for SQL Server, shifting machine learning workloads to Amazon SageMaker, or adjusting a small application for serverless computing on AWS Lambda.

7. Refactor (Re-architect) - When you make major changes to your resource to take advantage of cloud-native capabilities

The refactor or re-architect strategy involves changing an application's architecture to fully take advantage of cloud-native features and cloud-native capabilities.

You will often see this public cloud migration strategy applied to a legacy system with a monolithic design or outdated code that has become difficult to maintain. For example, a developer might redesign a monolithic application into microservices to improve scalability. Although this is often a more involved migration process requiring specialized cloud migration consulting, it typically offers greater long-term value.

When to choose each migration strategy

Type of cloud migration strategy	When to choose this strategy	Considerations
Retire	Retire a resource when it is barely used, with little impact on operations.	What new functionality will need to be introduced to cover the resource, if any?
Retain	Retain a resource when its compliance, dependencies, or operating costs make it critical to keep on-premises.	How will the resource interact with your cloud environment?
Rehost	Rehost a resource when it is already optimized and can run directly on existing cloud services.	Does the resource need changes for efficiency? If so, consider replatform or repurchase.
Relocate	Relocate a resource when it has an equivalent cloud structure.	An interconnected cloud structure requires careful governance to meet the same standards as your on-premise structure.
Repurchase	Repurchase a resource when there is an equivalent, updated, or more cost effective cloud licensing model.	Which cloud licensing structures or services have the least cost and greatest functionality with the closest operational capabilities to your existing resource?
Replatform	Replatform a resource when you want to take advantage of your cloud’s capabilities and you can do so without major rework of the resource.	How can you easily build cloud service connectors into your resource?
Refactor (Re-architect)	Refactor or re-architect a resource when you determine the ROI of modernization outweighs the cost to perform a major rework.	What is the timeframe and cost to refactor your resource vs retaining and refactoring at a later date?

The cloud migration process goes beyond current and target technologies to cover all facets of organizational change management. It includes the following three steps:

Step 1—Assess

Before conducting an on-premises to cloud migration, conduct a thorough cloud readiness assessment to have the best chance of transformational success. The AWS Cloud Adoption Framework (CAF) evaluates an organization's readiness across six perspectives: Business, People, Governance, Platform, Security, and Operations. The Assess phase is built on the principle that you can't effectively move what you do not measure. A comprehensive IT landscape assessment is required to baseline your current environment.

These cloud migration tools help discover specific readiness gaps. By working to close these gaps, you can improve the success rate of your cloud migration projects and overall cloud adoption. In some cases, organizations will find that it is not possible to migrate certain workloads or infrastructure to the cloud, where the decision will be made to retain or retire the component.

Perform a cloud readiness assessment

Organizations can use the AWS Cloud Readiness Assessment to self-evaluate their current readiness to adopt a cloud strategy. This tool provides insights into the readiness of your business, people, governance, platform, security, and operations. AWS also offers a free, detailed migration assessment in which experts score your organization across several verticals, including business value, operations, and security and compliance.

Perform an application portfolio analysis

An application portfolio assessment is the process of gathering application, infrastructure, and business context data to discover, analyze, and plan for moving the portfolio to the cloud. Performing an application portfolio analysis helps you to understand how to migrate and modernize your workloads. Complex infrastructure, integrations, and identity considerations often support existing applications. An application portfolio assessment helps identify legacy systems that may be too brittle for a rehost, as well as sensitive data that requires specific security measures. Identifying key stakeholders and data sources during this stage allows you to establish clear migration projects and priorities.

Organizations can use AWS Transform to discover their portfolio, by collecting information about their on-premises environment.

Step 2—Mobilize

The mobilize phase is the architectural phase of your migration.

Select target architectures and cloud migration tools

In the mobilize phase, you first select your target architectures and cloud migration tools. Choosing these in advance will guide how long the migration process will last.

Build a cloud migration roadmap

A cloud migration roadmap is a time-based planner to help you achieve your migration goals.

You will choose the best migration strategies for your current environment based on your business needs and your cloud providers’ target architecture. Key tasks include:

• Prioritization of migration resources.
• Development of migration targets, such as speed, downtime, and operational capacity.
• Staff training for migration and post-migration service management and operations.
• Creation of target architectures and services
• Test of the migration environment with smaller or less critical systems or services

Using an architectural decision record (ADR) during the mobilization phase is strongly recommended. This living document covers each architectural decision and its rationale.

For instance, many organizations choose to refactor or retain legacy software systems. Refactoring can be a lengthy and costly process, but it typically delivers ROI after several years. An ADR can help you determine whether a refactored structure, dependencies, interfaces, and construction techniques might provide improved performance.

Step 3—Migrate and modernize

The actual migration process itself depends on the size and requirements of your organizational shift. For smaller or simpler transitions, you can deploy standalone workloads individually using basic migration tools, as these projects do not require the development of repeatable framework patterns or automated templates. Larger or more complex enterprise migrations often require the deployment of standardized structural patterns that scale across multiple interconnected applications, underlying infrastructure components, distinct organizational units, and cross-functional projects.

AWS offers several cloud migration tools, such as AWS Transform and the AWS Database Migration Service, to support complex migrations. You can also consider building migration factories to handle larger-scale cloud migrations. A migration factory is a collection of concurrently operating teams, with specific teams conducting rehosting, re-platforming, and refactoring migrations. Refactoring teams typically work similarly to software development teams, with DevOps-style support.

Developing automation plans, processes, and tooling based on an initial migration helps accelerate cloud migration processes across the organization.

Overcoming cloud migration challenges

Successfully migrating to the cloud requires navigating a range of technical and organizational obstacles. Addressing these challenges early in your cloud migration strategy planning process will help you maintain operational consistency throughout.

Minimizing business disruption

Downtime is one of the primary considerations in planning the migration process. Minimizing business disruption during your cloud migration relies on the methods you use per resource. For example, to avoid downtimes of business-critical services, you can keep your on-site service running alongside a replicated cloud service until you are sure the two have performance parity. This approach maintains customer or user expectations for service delivery.

You can also minimize downtime in your IT environments by scheduling migration activities during low-demand times for each service. Phased approaches help by transitioning workloads incrementally.

Meeting data security and compliance requirements

Setting up cloud security measures in advance, such as account and service encryption and privileged access management, can help you mitigate the risk of unauthorized access before and during cloud migrations. Plan for robust security during transit of your data to the cloud, such as using IPSec tunnels or setting up a direct connection. Your choice of secure transfer methods will impact migration timelines. Your data backup strategy may need adjustment.

Meeting compliance requirements in highly regulated industries requires special consideration. Data residency and regulatory requirements dictate how and where you store information in a cloud environment. This can impact your choice of cloud location, as well as the path your data takes during cloud migration.

Maintaining business continuity

One outcome of a successful cloud migration is better integration with your organization’s business continuity goals. Although the migration itself involves shifting resources, the end goal is often to build a more resilient architecture that supports disaster recovery and higher availability using the native features of cloud-based infrastructure. By baselining performance before the move and validating success afterward, businesses can make sure that the cloud-based systems perform as expected without compromising long-term stability.

Mitigating vendor lock-in

Most cloud providers offer a range of services for migration to the cloud but it's also important to consider your ability to migrate out. The cloud vendor and set of services you choose for your cloud migration can affect your business operations and decision-making in the future. Avoiding vendor lock-in means making sure that you can port your applications and data out to another cloud or back on-site, if you need to. For example, self-managed instances on Amazon EC2 are highly portable to other cloud vendor environments. Your vendor should also provide tooling for multi-cloud management for cases where you need to combine services across multiple cloud and on-premises environments.

Optimizing resource management

A successful cloud migration depends on the people who will perform the migration and manage the cloud environment post-migration. In less cloud-aware teams, human error in migration and ongoing management can introduce inadvertent risks. You can use different resource management strategies to mitigate migration risks, based on the current cloud skillset and time availability of your IT team. Determine in advance whether you will need to train your team to adapt to new cloud technologies, recruit new team members, or outsource extra personnel or functionality during and post-migration.

With AWS cloud services, you can use your existing data and applications to scale and grow with AI and beyond from a strong foundational cloud provider. Selecting the right cloud migration strategies is critical to a successful migration overall. Whether you are considering cloud migration or are ready to start shifting your resources to the cloud, AWS provides guidance, assessment tools, and a range of partners to get started:

• AWS Transform is a collaborative enterprise IT transformation workbench powered by expert agents that accelerates cloud migration, application modernization, and continuous tech debt reduction. Your cloud migration journey can begin with AWS Transform.
• AWS Transform MGN is a dedicated rehosting capability that automates the conversion of your source servers (physical, virtual, or cloud) into native Amazon EC2 instances. AWS Transform MGN handles continuous block-level replication, automated machine conversion, and orchestrated cutover, enabling you to migrate workloads with minimal downtime and without modifying your source environment.
• The AWS Migration Acceleration Program (MAP) is a comprehensive and proven cloud migration program based on AWS's experience migrating thousands of enterprise customers to the cloud. MAP provides cloud tools, tailored training approaches and content, expertise from Partners in the AWS Partner Network, a global partner community, and AWS investment. It can reduce migration costs and technical debt. The AWS Migration Acceleration Program helps you build strong AWS cloud foundations, accelerate and reduce risk, and offset the initial cost of migrations.

Visit AWS Migration & Modernization to find your best-fit cloud solutions on AWS today.

FAQ

What is cloud migration consulting?

Cloud migration consulting helps you to gain a greater understanding of the scope, duration, and resources needed to migrate to the cloud. Cloud consultants are skilled professionals who can map your current architecture to cloud services, strategize about optimal configurations and cloud costs, and help walk you through the entire process. Hiring a consultant can be a useful guide in your business transformation.

How can you choose the right cloud provider?

Look for a cloud provider with a wide range of services to suit all architectures, with advanced AI and analytics services that connect to major AI models. The vendor will have purpose-built migration and export tools, support available on demand, and a list of partners to help with migration or new service delivery. A vendor should have an established presence in the field, in-region services, and positive customer experiences in digital transformation.

Is it necessary to have a separate data migration strategy?

A data migration strategy assesses your organization’s data storage, structures, and architectures, to determine how to migrate that data. A separate data migration strategy can help to modernize your organization’s data for consumption by AI and analytics, by applying architectural constructs such as a data lake or a data mesh.

Is it better to migrate over the internet or using devices?

Deciding whether to migrate your resources over the internet, or by delivering storage devices to physical cloud locations depends on many factors. These factors include time-to-migrate, security, cost, and complexity. You may choose a combination of internet-based and device-based transport for your migration, depending on your assessment of each resource.