⚡ dbt Fusion Upgrade Training Repo

This repo is a training project for Coalesce Upgrading to Fusion trainees to practice upgrading from dbt Core → dbt Fusion. It includes a sample dbt Core project plus a mix of hard and soft blockers that you’ll need to resolve in order to complete the upgrade.

The goal: gain hands-on experience troubleshooting and adapting real-world patterns during a Core → Fusion migration.

---

🚀 How to use this repo

1. Create your branch

   • Branch off main using the format:

     firstname_lastname_fusion_migration
     

2. Migrate the project

   • update the profiles.yml and/or the dbt_project.yml file to use your Snowflake ZNA creds
   • Make sure you run dbt seed in dbt Core first to create the underlying data sources
   • Update the repo from dbt Core to dbt Fusion.
     • If in VC Code, start by running the migration helper dbtf init --fusion-upgrade and follow the workflow
     • This will run the autofix package, anything left over will need to be converted by hand
   • Work through both the hard blockers (things that prevent Fusion from compiling/running until fixed) and soft blockers (warnings, best practices, or optional improvements).

3. Publish your work

   • Push your branch to GitHub.
   • Open a PR for review so others can see how you approached the migration.

---

🔒 Hard vs Soft Blockers

• Hard blockers: Must be fixed before Fusion will run (e.g., deprecated configs, broken macros).
• Soft blockers: Won’t prevent execution but highlight patterns that should be updated for maintainability or Fusion best practices (e.g., model naming, semantic configs).

---

📚 Why this repo exists

Migrating from dbt Core to dbt Fusion often involves resolving subtle differences in configuration, orchestration, and semantics. This training repo provides a safe sandbox to practice:

• Identifying blockers
• Applying migration strategies
• Gaining confidence before working with customer projects

---

👉 Ready to go? Clone the repo, create your branch, and start your migration journey.