Before learning about migration, first consider what user companies see as ideal core system. They likely want core system that keeps costs down, migrates stable current system safely, and supports new requirements. In other words, they want to keep strengths of current system, achieve stable migration, meet new requirements, and introduce it at low cost.

Four Points User Companies Request in Migration

• You want to make use of existing system assets.
• Stable system migration
• You want to reduce migration costs.
• Add new requirements and increase the benefits of implementation

Because migration is based on the current system, it comes with various constraints. As a result, it cannot always support many new business requirements. At the same time, companies naturally want to gain as much value as possible. Several migration methods exist to help strike this balance.

What migration methods are available? Let’s start by understanding the differences between rehosting, rewriting, and rebuilding. Read on for a clear overview.

Among migration methods, the terms rehosting, rewriting, rebuilding, and replatforming are often heard. Let’s look at the characteristics and differences of each.

Rehost

Rehosting is method that replaces IT foundation, including hardware and infrastructure environment, without modifying applications or programs.

This applies when hardware such as servers or storage is replaced due to the end of maintenance support or the need for higher specifications. Rehosting also includes so-called Lift & Shift, where an on-premises environment is moved to a cloud environment as is.

Migration rarely ends with rehosting alone. It is often carried out as first phase of large-scale migration project.

A similar term, “replacement,” is also used, but it is broader in scope. It may refer to rehosting, or to replacing an existing application with a new one.

Replatform

Replatforming is method that optimizes runtime foundation, such as OS, middleware, and databases, for new environment without major changes to application structure or business logic.

For example, this applies when migrating a system running in an on-premises environment to the cloud by replacing middleware or the DB with cloud-compatible options, or by optimizing the system configuration.

It requires slightly more modification than rehosting, but can improve performance and operability and optimize costs. On other hand, because business requirements themselves basically do not change, it offers less flexibility than rebuild.

Rewrite

Rewriting is an approach that rewrites the program to fit a new environment or technology while keeping the application's functions and business specifications unchanged.

For example, rewriting an existing program written in VB into a newer language such as Java or C# is considered a rewrite.

Compared with the rebuild approach discussed later, this can be achieved in a shorter time and at lower cost, but it is difficult to make major changes to business requirements, and flexibility is somewhat limited.

The term “conversion” is also commonly used, but it has a broad meaning. It may refer to rewriting itself, or to the broader conversion of programs, data, or file formats.

Rebuild

Rebuild is method that retires current system and reconstructs new system from ground up. It is basically same as scratch development, but some cases carry over business specifications from current system assets or migrate only data.

Because this approach starts again from requirements definition, it makes it easier to review and improve business requirements, making it the most flexible and future-ready option.

However, it is important to note that this approach requires a considerable project timeline and budget. In a broad sense, introducing a new ERP package system may also be considered a rebuild.

Summary

Rehost, replatform, rewrite, and rebuild can be organized by perspective of how much change is made.

• Rehosting: Changing only the environment, with minimal risk
• Replatform: Optimize Foundation and Migrate
• Rewrite: Modernizing technology while maintaining business operations
• Rebuild: Fully Renew Business Processes and Systems

Understanding these differences makes it easier to consider a migration strategy that fits your company.

Next, let’s look more closely at the key points to consider when moving forward with migration.

Let us revisit the four points that user companies seek in migration, as discussed earlier. Whether rehosting, rewriting, or rebuilding, no single approach can meet every requirement. Companies must carefully assess the trade-offs among their requirements and decide which approach to adopt or how to combine multiple approaches. When migrating a core system, the number of stakeholders involved becomes very large, and the difficulty of launching the project rises significantly.

Various issues can also arise when executing a project. Here, we will examine common problems and risks.

• The specifications and requirements of the current system are unclear.

One common issue with legacy systems is incomplete development and design documentation. In some cases, the people involved in building the current system are no longer available, and it is difficult to secure knowledgeable personnel, causing the system itself to become a "black box." This can be a major obstacle to making effective use of existing assets.

• Cost escalation

As the project progresses, incidents such as additional requirements and the need to reflect modifications to the current system may occur. Response costs may also increase due to external factors, such as OS upgrades or specification changes in peripheral systems. The longer the project continues, the more likely these issues are to arise.

• Resource Management Is Difficult

In large-scale projects, controlling human resources is especially difficult. In addition to securing personnel, work often becomes dependent on specific individuals such as project managers and subject matter experts, and workloads tend to concentrate in certain areas.

These factors often cause projects to be suspended or canceled. This should help clarify why migration projects are so challenging.

It is easy to describe migration as “moving an existing system to a new system,” but in practice it is a highly demanding initiative. Careful preparation and planning are essential, and it is especially important to clarify the specifications and business processes of the current system. One approach is to use system mining tools to understand the current system’s specifications and workflows. It is also important to consider building mechanisms that avoid relying on manpower alone, particularly in the testing phase, where workloads often expand significantly.