Our migration methodology

Leverage our experience and expertise from thousands of customer engagements. Feel confident with our methodology for proven success at every stage of your MongoDB migration.

Every migration is unique, and this methodology gives us a blueprint for developing a repeatable program that can provide you with confidence and predictability on your journey. While this methodology is not an explicit step-by-step process, they are the common phases we see most teams go through during successful migration projects.

Analyze

Before diving into designing a MongoDB solution, we first need to unpack the legacy architecture. Our Transformation Consultants specialize in providing tools, assessments, and workshops that help our customers understand how their current systems are connected, and how that can map to a future state MongoDB solution.

We start by establishing complexity scores for target applications by looking at the following considerations:

• Database structure and data dependencies
• Data-domain boundaries and business logic within stored procedures
• Service boundaries, such as business functionality, data access patterns, and scalability

Depending on where you are in your modernization journey, this analysis gives us a roadmap for which applications to prioritize, the feasibility of each, and the value to the business.

This information also informs migration tooling and approaches, shapes the plans to migrate those selected applications, and shows how to tackle more complex use cases that may span across multiple data domains.

At this stage, we also begin to identify what skills gaps exist within your development and operations teams, so we can later build an enablement plan ahead of a go-live date.

Design

Before migrating data, we need to design the data model to support the most common and critical access patterns. We also consider the complete end-to-end architecture and make recommendations where simplification can occur.

This stage is critical because data modeling with MongoDB’s document model is fundamentally different from relational models. When designed correctly, the gains to business agility and developer productivity are massive. An effective MongoDB schema establishes data as a product within the organization, allowing other teams and applications to more easily consume data in the future.

MongoDB Relational Migrator, our preferred tool for supporting these migrations, makes it easy to develop an effective schema through recommendations and a visual mapping experience.

Refactor

Data and databases aren’t the only part of a migration project — applications need to be updated too.

Relational databases need their embedded business logic within stored procedures to be translated to the appropriate place in the future technology stack, whether it’s shifting logic to the application layer, taking advantage of the powerful aggregation framework, or creating server-side functions with MongoDB Atlas.

Our team can provide advisory support during your refactoring efforts. But for teams who need hands-on-keyboard assistance, we can provide you with a complete development team in partnership with certified MongoDB application development partners to execute this work for you.

Migrate

Now we’re ready to migrate data from the existing relational databases to their MongoDB destination.

We start with selecting the appropriate data migration strategy, migration tooling (such as MongoDB Relational Migrator), and cutover strategy given the database’s needs (like data size) and the business’s requirements (like acceptable downtime).

The migration plan is often tested against a representative lower environment, and a dry-run migration is performed. Once we’re confident in our migration plan, cutover begins, and our team provides support through the entire experience.

Operate

After the cutover is complete, we run a post-migration health check to validate the integrity of the migrated data and identify opportunities to optimize performance — which is especially important when migrating to a cloud platform like MongoDB Atlas for the first time.

We also put the management tooling and operations playbooks in place to help you effectively operate MongoDB going forward. Business-critical applications require a robust operational model that considers:

• Backup and disaster recovery strategy
• Security
• Monitoring and alerts
• Scalability

Finally, we conduct training and provide enablement plans to upskill your teams on MongoDB features and best practices so they can be successful and confident developing with and managing the platform.

While this is the end of this methodology for migrating one application, it’s not the end of your broader modernization journey. We use your first successes to begin designing the next application to migrate, bringing you closer to your future-state solution built with MongoDB.