Discovering the myths in cloud migration
Cloud migration moves a current workload, including compute, storage, data, services, system integrations, and configuration, from on-premises to the cloud. The core goal of this process is to achieve particular business objectives. These business objectives are measurable from the perspectives of efficiency and effectiveness. Therefore, it is essential to plan this process with suitable assessment and metrics, design principles, strategy patterns, and best practices, and the right people, tools, and technologies.
In addition, we also need to identify the priority, risk, and relationship between the business goal and migration strategy. Finally, we need to focus on data and the results of the migration process. However, often organizations get distracted by noise and myths. This section will clarify certain myths and unfair practices about the cloud migration process. These will be presented as quotations in each subsection, followed by an explanation of the myth. This section aims to help you understand what not to do in a cloud migration process.
Myth – re-hosting a workload without preparing for the cloud
As we all know, the most common challenge of on-premises is that it is difficult to scale up and down with a change in demand. The cloud consistently demonstrates excellent performance for scalability. However, just re-hosting workloads does not ensure scalability. The Re-Architect pattern or cloud modernization are essential to ensure scalability. In most cases, applications are modernized through containerization and integrating cloud readiness by design. However, deploying an application on a container platform is not enough for scalability. We need to design the network infrastructure and configuration for the container platform to achieve the goal.
To avoid this problem, a cloud maturity model is beneficial. A cloud maturity model defines workload classification and determines which workloads need to be redesigned to make them cloud-ready.
Myth – relocating a workload without a cloud-native design
Relocating workloads to the cloud ensures HA by design. Enterprise applications demand HA for each cloud service to keep up with high demand levels. However, this is not necessarily true. A cloud provider will provide a certain level of service as defined in its service-level agreement (SLA). However, most of the time, it does not ensure business continuity. Therefore, enterprises need to ensure that business continuity and disaster recovery are in the plan, and modernization or Re-Architect are mandatory patterns for a business-critical or production workload.
Myth – limited knowledge about cloud computing
A public cloud is not secure.
Security combines technology, tools, processes, practices, and cultures. In addition, appropriate implementation of regulation and compliance to avoid risks is essential in the cloud. Security can be ensured if it is the heart of development. Along with designing a secured infrastructure, developers carry a massive responsibility to secure an application. Although industries are still heading toward hybrid cloud solutions for their workloads, the cloud adoption by different industries, such as the financial and healthcare sectors, is also increasing. Therefore, the public cloud cannot blamed entirely for security issues when security is a shared responsibility including integrated processes, rightsizing, tools, methods, and practices.
Myth – cloud migration without an objective
Cloud migration is a technical strategy rather than a business strategy.
Cloud deployment directly impacts the revenue and business outcome for organizations. Therefore, it is essential to have a business objective for cloud migration. For any cloud migration project, there should be a proper business case analysis that will drive the migration project from the financial and technical points of view. As cloud migration projects are usually expensive, without the proper business case, there are chances the project might fail. The business case also helps to take correct technical strategy and decisions for the migration project and cloud platform. If an organization simply tries replacing an on-premises data center with the cloud as their platform of choice without any solid business case, it might end up as a waste of money and effort in the end. So we need to keep in mind that cloud migration should be driven by the business strategy, which could be to reduce the time to market, or allow for massive scaling up in case of increased business popularity.
Myth – limited or no assessment for cloud migration planning
We have a simple Excel sheet, which is enough for our workload analysis for cloud migration.
We need to remember that cloud migration is only beneficial if it is done right. Therefore, it is essential to plan the cloud journey properly with details.
Myth – inefficient assessment for cloud migration
One size can fit all.
The recent trend among enterprises is to choose a single cloud as their strategic cloud and put everything in that same cloud. The outcome is already visible – often, this is not the right decision. In addition to that, no single cloud migration pattern is suitable for every type of workload. Migration patterns vary for different technology/applications.
The landscape and business goals of the transformation need to be defined clearly for a successful cloud migration project. Often, enterprises prefer to use Re-Host as their only cloud migration pattern instead of Re-Architect. However, this is not always true. A proper analysis is essential to identify the correct cloud migration pattern for different workloads. In reality, 84% of digital transformation projects fail due to siloed data and unreliable integration approaches.
Myth – cloud migration versus cloud modernization
Move first, then innovate.
Monolithic legacy applications only focus on business logic and functionality. For example, developers often use files at a specific location to store the configuration. The main focus for developers is to implement the functionalities; unit tests, integration tests, and regression tests focus on business logic only. The data, integration, management, and DevOps are often a secondary focus for developers. Therefore, moving applications as is usually makes the management more expensive and cumbersome. A more holistic focus on configuration, management, DevOps, data, and integration can modernize an application and make it suitable for cloud migration.