Database Integration

Overview

Once your connections are defined (see the Connections guide), Framework makes it easy to run database queries, execute statements, and tap into the internal framework.db. Use the high-level helpers to avoid hand-managing DBI connections.

Running Queries

Get Data

# Query and return results
users <- db_query("SELECT * FROM users WHERE active = true", "analytics")

Execute Statements

# Execute without returning data
db_execute("UPDATE stats SET last_run = NOW()", "analytics")

Manual Connections

Need direct DBI access? Grab a connection and clean up afterward:

conn <- db_connect("analytics")
DBI::dbListTables(conn)
DBI::dbDisconnect(conn)

Prefer db_query() and db_execute() when possible—they open and close connections automatically.

Supported Databases

Framework supports the following database drivers out of the box:

Driver	Package	Use Case
postgres / postgresql	RPostgres	PostgreSQL databases
sqlite	RSQLite	Local SQLite files
duckdb	duckdb	Analytical workloads, Parquet files
mysql / mariadb	RMariaDB	MySQL/MariaDB
sqlserver / mssql	odbc	SQL Server

Object storage connectors (S3-compatible endpoints) are covered in the Connections guide.

DuckDB Configuration

DuckDB supports additional options in settings.yml:

connections:
  warehouse:
    driver: duckdb
    database: data/warehouse.duckdb
    read_only: false
    memory_limit: "4GB"
    threads: 4

The Framework Database

Every project includes a local SQLite database (framework.db) for internal tracking:

cache_entries <- db_query("SELECT * FROM cache", "framework")

This tracks:

• Data integrity hashes
• Cache metadata
• Results registry

Use regular SQL to inspect or export these tables, just like any other connection.