Overview of the buster deploy Command

The buster deploy command publishes your Buster project to the platform, making your semantic models available for data analysis.

Basic Usage

To deploy your Buster project:

buster deploy

This command:

  1. Validates your buster.yml configuration
  2. Verifies your semantic model definitions
  3. Publishes your models to the Buster platform
  4. Makes them available for natural language querying

Deployment Process

When you run buster deploy, the following happens:

  1. Configuration Validation: Checks your buster.yml file for correctness
  2. Model Validation: Verifies that all semantic models are properly structured
  3. Data Source Verification: Confirms that the configured data sources are accessible
  4. Model Publishing: Uploads your semantic model definitions to the Buster platform
  5. Indexing: Processes your models to enable natural language understanding

Command Options

The following options are available:

  • --project-file FILE: Specify a custom buster.yml file location
  • --dry-run: Validate your project without actually deploying
  • --verbose: Show detailed information during deployment
  • --force: Force deployment even if there are minor validation issues

Example

Deploy a project with the default configuration:

buster deploy

Deploy a project with a specific configuration file:

buster deploy --project-file ./configs/prod-buster.yml

Deployment Lifecycle

Your project goes through several stages during deployment:

  1. Validation: Initial checks of your configuration and model definitions
  2. Compilation: Processing your semantic models into Buster’s internal format
  3. Publishing: Making your models available on the platform
  4. Activation: Enabling your models for querying

Troubleshooting Deployment

If you encounter issues during deployment:

  1. Validation Errors: Check your buster.yml and semantic model files for syntax errors
  2. Connection Issues: Verify that your data source connection details are correct
  3. Permission Problems: Ensure the database user has the necessary permissions
  4. Timeout Errors: For large projects, try increasing the deployment timeout

Best Practices

  1. Test Before Production: Use the --dry-run flag to validate your project before actual deployment
  2. Regular Deployments: Deploy frequently as you make changes to keep your models up to date
  3. Versioning: Use git or another version control system to track changes to your semantic models
  4. Validation: Run buster validate before deployment to catch issues early

Next Steps

After deploying your project:

  1. Test your models with natural language queries
  2. Monitor query performance and accuracy
  3. Refine your semantic models based on usage patterns
  4. Add more detailed metrics and relationships as needed