v1.0.21 - Latest Bundle with Simplified Cluster Management
What's Changed
π Latest Bundle Configuration
- Simplified cluster management: Let bundle handle cluster creation and naming
- Fixed app startup: Added app.yaml and improved Streamlit app
- Enhanced debugging: Better cluster visibility and status checks
π Current Bundle Features
- Cluster configuration: Spark 16.3, m5d.large, On-Demand instances
- App configuration: Streamlit app with proper command and error handling
- Job integration: Databricks job that can be triggered from the app
- Bundle deployment: Automated deployment with proper resource management
π§ Technical Details
- Uses Warning: unknown field: apps
at targets.dev.resources
in databricks.yml:20:7
Warning: unknown field: apps
at resources
in resources/app.yml:2:3
Warning: unknown field: clusters
at resources
in resources/clusters.yml:2:3
Error: cannot resolve bundle auth configuration: validate: more than one authorization method configured: oauth and pat. Config: host=https://ias-dataplatform-dev.cloud.databricks.com/, token=, client_id=your-client-id, client_secret=. Env: DATABRICKS_HOST, DATABRICKS_TOKEN, DATABRICKS_CLIENT_ID, DATABRICKS_CLIENT_SECRET for cluster lifecycle
- 120-second wait for cluster readiness
- Proper app.yaml configuration for Streamlit
- Enhanced error handling in app.py
- OAuth authentication throughout
π‘ Expected Results
This release should:
- β Create cluster with correct name - hello-world-cluster
- β Start Streamlit app properly - No more python app.py errors
- β Provide clear debugging - Better visibility into deployment process
- β Handle all edge cases - Robust error handling and recovery
Monitor the deployment in GitHub Actions and Databricks UI for the hello-world-cluster.