Contact Us 1-800-596-4880
  1. Homepage
  2. Mule Runtime (4.8)
  3. Mule 4 for Mule 3 Users

  4. Migration to Mule 4

Migration to Mule 4

For your Mule 3 apps to take advantage of Mule 4, migrate them to Mule 4 through one of the following processes:

  • A fully manual migration of connectors, modules, transports, expressions, and API gateways to their Mule 4 counterparts.

    For guidance with a fully manual migration, see Prepare to Migrate and review Manual Migration Process.

  • A partially automated migration with the Mule Migration Assistant (MMA), which is available as an open source project on GitHub.

    For guidance using MMA to assist in a migration, see the user documentation (on GitHub). MuleSoft suggests that you try MMA to determine whether you prefer it to a fully manual migration. Note that you can run MMA on a sample application by following steps in the Migration Tutorial (on GitHub).

Important

Mule Migration Assistant (MMA) is subject to the terms and conditions described for Community connectors. Additionally, Mule Migration Assistant is distributed under the terms of the 3-Clause BSD License.

A benefit of using MMA is that it automates part of the migration of Mule 3 apps to Mule 4. Though the MMA does not migrate the app entirely, it provides significant help with the following tasks:

  • Migrating the project structure.

  • Automatically creating descriptor files, such as pom.xml or mule-artifact.json.

  • Automatically performing many adaptations of the app’s code.

  • Providing guidance on how to manually migrate components and patterns that cannot be migrated automatically.

MMA helps with such tasks, but it does not deliver a fully functional app. Some additional steps are required. Those steps include testing and can also include:

  • Manual migration of unsupported components.

  • Manual migration of complex MEL expressions to DataWeave that cannot be automatically converted by MMA.

  • Manual migration of complex DataWeave transformations that cannot be automatically converted by MMA.

  • Prior migration of the DataMapper with the DataWeave Migration tool, before running the MMA.

It is also important to note that the MMA does not support incremental migrations.