What is cloud migration?

Cloud migration explained

NetApp offers a simple definition for cloud migration, stating that “a cloud migration is when a company moves some or all of its data center capabilities into the cloud, usually to run on the cloud-based infrastructure provided by a public cloud service provider such as AWS, Google Cloud, or Microsoft Azure”.¹

Techopedia defines cloud migration or a business progress outsourcing (BPO) as “the process of partially or completely deploying an organization’s digital assets, [platform] services [and similar organizational infrastructure], IT resources or applications to the cloud”.² Techopedia continues noting that the enterprise’s migrated assets are usually accessible behind the cloud’s firewall and that an enterprise’s cloud migration process “often involves merging an on-site IT infrastructure with a hybrid cloud solution [that] can be accessed [online, and that] Hybrid cloud solutions transition between one or more cloud providers and usually provide on-demand and provisioned server space, applications and services”.²

Cloud migration strategies

To help your enterprise with preparing for their cloud migration, it’s best to develop a cloud migration strategy that works best for your needs and incorporates both your infrastructure components and workload migration options into it. Put simply, if you have a comprehensive view of your infrastructure and options for migrating each workload, then it will make the process of planning and performing your cloud migration much easier.

“The 6 R’s of cloud migration” are a series of cloud migration strategies that can help you determine which cloud migration strategy is best for your enterprise:

• Rehost
• Replatform
• Refactor
• Repurchase
• Retain
• Refactor

Rehost 

Also known as the lift and shift model or lift-and-shift, rehosting is arguably the most straightforward cloud migration strategy. It involves transitioning apps, virtual machines, and server operating systems from their current hosting environment onto a public cloud infrastructure without any changes. 

Rehosting works well as an entry point for large optimization projects, particularly when enterprises are looking to quickly transition their on-premises infrastructure. For example, if your enterprise’s data center is having issues, you could opt to rehost that data center’s workloads onto the cloud. An advantage of rehosting is that it’s a straightforward process, while a disadvantage is its limited use of cloud efficiencies. Simply rehosting workloads to the cloud doesn’t immediately yield the benefits of cloud-native features.   

Replatform 

Replatforming or lift and optimize is a more complicated cloud migration strategy than rehosting, as it often involves the optimization or the upgrading of apps, and middleware. 

Replatforming exists somewhere between a lift and shift and a supercharged application re-architecting.   

Replatforming helps enterprises to make the most of their cloud benefits, often reworking their sourcing environment and making it more cloud compatible, improving app functionality, and side-stepping post-migration work.

Refactor

Sometimes called rearchitecting, refactoring is defined by Margaret Rouse as “the process of altering an application’s source code without changing its external behavior”.³ A potential opposite of the lift and shift cloud migration strategy, refactoring is prompted by a great desire for improvement and for achieving a particular goal or target. Rouse notes that “the purpose of re factoring is to improve some of the nonfunctional properties of the code, such as readability, complexity, maintainability and extensibility”.³

There are three types of refactoring: 

• Code refactoring – Synonymous with general term “refactoring”.
• Database refactoring – Making basic changes to a database schema to improve its design without altering informational and behavior semantics. 
• User interface (UI) refactoring – Making basic changes to the UI without altering functionality.

These features help refactoring to be a cost saving strategy with an added benefit of potentially “[extending] the life of source code, preventing it from becoming legacy code”.³

Repurchase

Repurchase or drop-and-shop is a cloud migration strategy that involves swapping the proprietary app in-use for a new cloud-based platform or service. Enterprises can replace your app with a different version or product and the newer app should offer greater business value than the previous or otherwise existing, on-premises app, including features such as broad accessibility, freedom from any infrastructure to maintain, and pay-as-you-go pricing models. 

The author of the AWS Prescriptive Guidance page notes that “repurchasing the application typically reduces costs associated with maintenance, infrastructure, and licensing”.⁴

Common use cases for a repurchasing migration strategy include the following examples:

• Moving from a traditional license to SaaS
• Version upgrades or third-party equivalents
• Replacing a custom application⁴
  

Retain

Retain or hybrid model is a cloud migration strategy that’s leveraged to retain certain components of your IT infrastructure, such as your source environment, apps, workloads, or databases, on your legacy infrastructure. With this cloud migration strategy, your enterprise might have some workloads hosted on the cloud and some workloads retained on-site.   

Enterprises might choose not to migrate these infrastructure components for several reasons, including that they’re not ready for migration or there are security or vulnerability issues.  

The AWS author offers several use cases for why your enterprise might choose to retain applications and similar IT infrastructure:  

• Security and compliance – Remaining in compliance with data residency requirements.
• High risk – An application requires a detailed assessment and plan prior to migration.
• Dependencies – You need to migrate one or more other applications first.
• Recently upgraded applications – Postpone migrating the application until the next technical refresh because you recently invested in upgrading your current system.
• No business value to migrate – Some applications only have a few internal users and aren’t worth the cost of migrating them onto the cloud.
• Plans to migrate to SaaS – Waiting until the SaaS version is released by the vendor. This is a common strategy for vendor-based applications.
• Unresolved physical dependencies – An application is dependent on specialized hardware and doesn’t have a cloud equivalent, such as machines in a manufacturing plant.
• Mainframe or mid-range applications and non-x86 Unix applications – Applications require careful assessment and planning before migrating them onto the cloud. Some examples of mid-range applications include IBM AS/400 and Oracle Solaris.
• Performance – Retaining applications based on their performance. For example, keeping poor performing [known as “zombie”] applications or applications that have an average CPU and memory usage between 5 and 20% over a period of 90 days [known as “idle”] applications in your source environment.⁴

Retiring

Retiring is the primary cloud migration strategy for apps or environments that need to be archived or decommissioned, shutting off these infrastructure components without decreasing productivity or value, and replacing their functionalities through other services and components. The AWS author notes that “retiring [an] application means that you can shut down the servers within that application stack”.⁴

Common use cases for leveraging the retiring strategy for infrastructure components are very similar to those for the retaining strategy, and include the following examples:

• No business value in retaining or moving the app to the cloud.
• Eliminates the maintenance and hosting costs.
• Reduces or eliminates security risks, especially if this involves legacy systems or components that are no longer supported.
• Zombie apps or idle apps. 
• No inbound connection to the app for the last 90 days.

Benefits of cloud migration

There are many different reasons why an enterprise would want to undergo a cloud migration. Enterprises often migrate to the cloud step away from their older legacy infrastructure, such as aging servers or software or hardware solutions that are performing poorly, helping an enterprise to improve their performance and efficiency.   

Margaret Rouse argues that “cloud computing is attractive [due] to its scalability, ease of management and low costs [and] cloud migration facilitates the adoption of flexible cloud computing”.² 

In Microsoft’s eponymous article about cloud migration’s benefits, the author notes that although cloud migration has several challenges including planning and initial cost, it can include the following general benefits:  

• Backup, recovery, and failover: Several cloud providers help to improve business continuity by offering built-in, one-click backup and recovery capabilities, and offer the ability to store backups in different geographic regions.
• Compliance: Several cloud platforms include specialized offerings to seamlessly meet compliance requirements, which is an asset for regulated industries like finance, healthcare, and government.
• Flexibility and scalability: 
• Enterprises can opt to only pay for the cloud resources they use and can direct resources as needed.
• The hybrid cloud model offers flexibility for enterprises looking to continue getting value from certain existing IT investments (aka workloads) and migrate other investments in need of optimization.
• Enterprises can improve performance by running workloads closer to where their end users or customers are located.
• Enhanced security: Built-in security features and specialized cloud security tools help keep enterprises’ resources protected with additional automatic security patching offered by the cloud provider.  
• Optimized costs: Cloud migrations often result in cost savings from resources, maintenance, and real estate [from telecommuting], and the ability to optimize workloads that previously cost more to run on the legacy infrastructures. 
• Simplified management and monitoring: Central management tools help enterprises to monitor and manage their on-premises datacenters and cloud resources from a single application.⁵
  

Yifat Perry states that cloud migration can improve performance, noting that “applications and websites hosted in the cloud can easily scale to serve more users or higher throughput, and can run in geographical locations near to end-users, to reduce network latency”.¹  Describing how a cloud migration offers a better digital experience for end users, Perry writes, “Users can access cloud services and data from anywhere, whether they are employees or customers. This contributes to digital transformation, enables an improved experience for customers, and provides employees with modern, flexible tools”.¹

Why is migrating to the cloud a complex process?  

Before embarking on a cloud migration journey, it’s imperative to understand that the cloud migration process is not a simple one and it requires careful analysis, planning and implementation. Cloud migration can seem like an easy decision for enterprises to rush towards, but before they begin their journey to the cloud, they should be sure to understand the challenges to cloud migration along with its benefits.  

Financial cost

Following a cloud migration, an enterprise may experience periods of stark improvement with streamlined processes, increased innovation and efficiency, and reduced costs. The siren call of these benefits is enough to entice any enterprise into opting for cloud migration without considering the financial costs of implementing the whole process. Enterprises looking to migrate to the cloud to reduce the costs associated with their IT infrastructure may quickly discover that the cost of their migration is greater than the cost of maintaining their on-site data center.  

Financial concerns cover almost every aspect of the cloud migration and extend from the immediate cost of the actual lift and shift to the longer-term risks of slow or low adoption. Anyone who fails to properly research and plan their migration can easily be surprised at the cost during or after their migration. “If an organization lacks visibility into its IT infrastructure and how it is used, then consumption of cloud services – and their associated costs – may exceed expectations”.⁶

A longer cloud migration process will likely cost more than a shorter one. For example, if application architecture for the cloud needs to be rewritten, then the cloud migration’s cost will likely increase.  Other examples that can increase cost include:  

• hiring skilled people and buying the tools required for a successful migration
• post-adoption training for users on the new systems
• bandwidth costs
• performance issues: 
• downtime 
• latency
• interoperability
• dependencies on non-cloud apps

Calculating the actual cost of a successful cloud migration for your enterprise is critical to success and can prevent the exercise from becoming a cost-sinking quagmire.  

Adoption resistance

People can pose the biggest challenge to a cloud migration’s success. Whether it’s pushback to begin the journey to the cloud for any number of reasons (fear of change or disruption, resistance towards learning a new system, etc.), the human element can stop a cloud migration before it starts. 

Convincing everyone at your enterprise to support a cloud migration can be a herculean, time-consuming task. Employees may be against the change for purely emotional reasons. They might also have a fundamental lack of understanding about the process or exaggerated concerns about potential disruptions.  

Shortage of cloud skills   

As mentioned in the previous section, people can be the biggest obstacle for a successful cloud migration. If an enterprise’s employees lack the necessary skills and knowledge to effectively manage and implement a cloud migration, then cloud migration may result in prolonged downtime and increased costs, and ultimately it may not be successful.  

An enterprise can opt to provide continuing education to help cultivate the latest cloud skills and strategies among their employees and offer cloud migration training before the migration process begins. Enterprises can also hire cloud migration consultants to help their employees to prepare and to improve their chances of an uncomplicated and efficient, successful cloud migration. 

Lack of a strategy and poor planning

For an enterprise to maximize the success of their cloud migration, they should carefully design and plan the cloud migration process and ensure that they’ve identified the migration strategies that best meet their needs. “Public, private, and hybrid cloud deployments all have their pros and cons, and different cloud providers and cloud distribution models (SaaS, PaaS, IaaS, etc.) are optimized for different use cases”.⁶

A coherent, detailed cloud migration strategy helps prevent enterprises from being overwhelmed by the cloud migration process. “A cloud migration strategy should describe the goals that the cloud migration is intended to achieve and how to measure success. Based on these goals, cloud services can be selected, and a strategy can be developed to guide the migration”.⁶

Complex architecture, scope creep, legacy applications, and a long migration process

During a cloud migration, the process of lifting and shifting data storages and applications from an enterprise’s on-site environments onto their cloud infrastructure is frequently performed in stages. Implementing the process in stages instead of all at once helps ensure that the most recent stage was successful before beginning onto the next stage of the process. 

When an enterprise has complex existing architecture, it increases the difficulty of planning and implementing a smooth cloud migration. “Certain systems may need to be co-located to provide high-quality services, and a complex architecture may make it difficult to identify and document interdependencies and develop a phased strategy for moving certain components or systems to the cloud”.⁶

An enterprise’s existing architecture doesn’t have to be complicated to create issues during their cloud migration process. Cloud environments are fundamentally different from on-site environments. Having employees with knowledge and skills for each environment increases the chances of a successful migration. The employees performing the cloud migration need to understand the type(s) of cloud architecture being leveraged and how to optimize its configuration. There are differences between migrating to a public cloud vs. a private cloud, or a hybrid cloud vs. multicloud. If the data and applications aren’t properly configured for the cloud environments it can result in complications, including performance issues and the inability to access necessary components or data. 

Lifting and shifting legacy applications can be problematic, especially if the legacy apps aren’t compatible or are too complicated for an enterprise’s cloud services. Apps that are too old or require dramatic changes for them to work in a new environment can result in delays or increased costs for the cloud migration process. 

The process of migrating data and applications from where they’ve been for years to a new environment can lead to scope creep, “the tendency to add more and more features to the project and can result in a never-ending cycle of work”.⁷ Scope creep can require more time and resources than were initially planned, and result in increased spending, slowdown, and delays, which in-turn can lead to the cloud migration going overschedule and result in budget overruns.

The longer it takes to complete the cloud migration, the greater the cost and impact the migration will have on its enterprise.  “During the intermediate stages of the cloud migration process, infrastructure costs are increased because the business is paying for both on-prem and cloud infrastructure”.⁶

The cloud migration process lacks ownership, fallback plans, and processes

Having a single point of contact that’s responsible for decision making and overseeing the whole cloud migration process can be a huge benefit. The entirety of the cloud migration is a complex affair, and there’s always a risk of miscommunication and confusion which can lead to additional costs, slowdown and delays.  

Having fail-safe plans i can help to both reduce downtime, providing a means for operations to quickly resume. Similarly, maintaining persistent uptime during the journey to the cloud is an arduous task. Incorporating fallback planning into your migration helps to reduce the risk of slowdown and delays, and additional costs.  

Data security and compliance, replication, and infrastructure as code

Data security and regulatory compliance are major concerns for every IT team. If your enterprise is embarking on their journey to the cloud, then security and compliance will certainly be a factor in your enterprise’s migration. While some cloud deployment strategies are better suited for adherence to data security and regulatory compliance, “data and applications hosted on the cloud must be secured at the same level as they are on-prem”.⁶

If falls on your enterprise to verify that “their cloud environments comply with applicable data regulations or laws (e.g., GDPR and HIPAA) and maintain high data security standards to protect against threats like data breaches or unauthorized access”.⁷ The process of maintaining this level of compliance can demand lots of resources because cloud storage systems can have different security protocols and architectures than on-site environments. Alternatively, cloud environments often have comprehensive security features that can help comply with security and compliance regulations.  

It is imperative that you backup your critical data – both as a safety measure during your cloud migration and to help increase your cyber resilience. Consider storing your data in multiple locations to help mitigate the risk of disruption and, if disruption does occur, to support disaster recovery efforts and an expedited recovery time objective (RTO). 

 “Adopting an infrastructure as code (IaC) approach for cloud computing can provide several benefits, such as faster and simpler deployments, improved scalability, and reduced costs”.⁷ If you do opt for leveraging IaC, then work to equip your enterprises IT teams with the skills and tools they need to both properly “implement IaC and to be comfortable shifting to a cloud-driven development process”.⁷