​
Oxy sync

The oxy sync command collects semantic information from your databases and stores it in your local repository for use by agents and other Oxy features. This semantic information includes table structures, relationships, and sample data that help agents understand your data model.

​
Command Overview

oxy sync [DATABASE_NAME] [OPTIONS]

​
Arguments

  • DATABASE_NAME (optional): The name of a specific database to sync, as defined in your config.yml file. If omitted, all databases will be synced.

​
Options

  • -d, --datasets <DATASETS>: Specify one or more datasets to sync within the database. Can be used multiple times.
  • -o, --overwrite: Overwrite existing files during synchronization. By default, existing files are skipped.

​
Usage Examples

​
Sync All Databases

oxy sync

This command will synchronize all databases defined in your config.yml file. By default, it will skip any files that already exist.

​
Sync a Specific Database

oxy sync my_database

This syncs only the database named “my_database” from your configuration.

​
Sync Specific Datasets

oxy sync my_database -d customers -d orders

This will sync only the “customers” and “orders” datasets from the “my_database” database.

​
Overwrite Existing Files

oxy sync -o

By using the -o flag, existing semantic files will be overwritten during synchronization.

​
Output and Warnings

The sync command provides feedback about the synchronization process:

  1. Skipped Files: When a file already exists and overwrite mode is off (default), the command will display a warning indicating which files were skipped.

  2. Overwritten Files: When the -o flag is used, the command will display a warning showing which existing files were overwritten.

​
Understanding the Synced Data

When you run the oxy sync command, Oxy will:

  1. Connect to each configured database
  2. Extract schema information, including tables, columns, relationships, and data types
  3. Generate semantic models representing this data
  4. Save these models as files in your project directory

These semantic files are then used by agents when you ask questions about your data, enabling them to generate accurate SQL queries and understand your data structure.

​
Configuration

Before using the sync command, ensure your databases are properly configured in your config.yml file. For example:

databases:
  - name: my_postgres_db
    type: postgres
    host: localhost
    port: 5432
    user: postgres
    password: mysecretpassword
    database: mydb

  - name: my_bigquery_db
    type: bigquery
    key_path: path/to/keyfile.json
    datasets:
      analytics: ["*"]
      marketing: ["users", "campaigns"]

See the configuration documentation for more details on setting up your database connections.