Supported Drivers & Native Engines — CATable Documentation

High-Speed Native Engine Implementations

CATable avoids slow generic JDBC or ODBC abstraction layers. Each supported engine communicates through specialized native Rust drivers for microsecond query dispatch.

1. PostgreSQL Native Driver

Built on high-performance async protocol bindings:

Native streaming of 100,000+ rows with constant memory footprint.
First-class support for PostgreSQL specific data types: UUID, JSONB, HSTORE, INET, and PostGIS geometries.
Direct visual inspection of Active Locks (pg_stat_activity) and Table Bloat.

2. SQLite with WAL Optimization

Integrated native SQLite engine with full extension support:

Automatic detection and monitoring of Write-Ahead Logging (WAL) files.
Support for SQLite extensions: sqlite-vec for vector embeddings, full-text search (FTS5), and JSON1.
One-click database optimization: PRAGMA optimize, VACUUM, and index integrity checks.

3. DuckDB Columnar OLAP Engine

Perform instant analytical queries directly against raw files without loading them into a database server:

-- Query multiple Parquet files directly in DuckDB
SELECT 
    date_trunc('month', transaction_date) AS month,
    sum(amount) AS total_revenue,
    avg(processing_time_ms) AS avg_latency
FROM 's3://analytics-bucket/logs/*.parquet'
GROUP BY 1
ORDER BY 1 DESC;

4. Redis Key-Value & Memory Inspector

RESP3 protocol driver for real-time cache telemetry:

Safe cursor-based key scanning via SCAN (never running blocking KEYS *).
TTL countdown timers, memory footprint analysis per key, and one-click key expiration modification.