Mar 15
Loading video player...
In this tutorial I show you how to build a complete CI/CD pipeline for Databricks using Databricks Asset Bundles (DABs) and GitHub Actions with OIDC authentication — no static tokens, no manual deployments, no drift between environments. Starting from scratch, we configure a service principal with minimum Unity Catalog permissions, set up OIDC federation policies so GitHub Actions authenticates without PATs, and structure a DAB that declares every resource as code: Spark Declarative Pipelines, jobs, schemas, volumes, an AI/BI dashboard, and a Genie Space. We then wire everything together with a GitHub Actions workflow that runs bundle validate → bundle plan → bundle deploy automatically on every push — across isolated dev and prod targets. What you'll learn: 1. How to structure a repo with Databricks Asset Bundles 2. How to configure dev and prod targets in databricks.yml 3. How to set up OIDC federation between GitHub Actions and Databricks — no PATs 4. How to use GitHub Environments to isolate secrets per environment 5. How to run CI/CD with bundle validate, plan, and deploy 🔗 GitHub repo: https://github.com/alessandro9110/Speech-To-Text-With-Databricks 📄 Medium article: https://medium.com/towards-data-engineering/ci-cd-on-databricks-with-asset-bundles-and-github-actions-9d6fc0a337e9