Remote Sources

You can query data that lives on the web or cloud storage.

HTTP / HTTPS

You can query public URLs directly using DuckDB’s read_* functions or by simply treating the URL as a table.

-- Auto-detect format
SELECT * FROM 'https://raw.githubusercontent.com/duckdb/duckdb/master/data/parquet-testing/glob.parquet';

-- Explicit function
SELECT * FROM read_csv('https://example.com/data.csv');

CORS Issues

Browsers block requests to servers that don’t send Cross-Origin Resource Sharing (CORS) headers. If a query fails due to CORS:

1. Enable the CORS Proxy in Settings.
2. PondPilot will route the request through a proxy server to bypass the restriction.

Cloud Storage (S3)

PondPilot supports reading from S3-compatible storage (AWS S3, MinIO, GCS, R2).

Public Buckets

SELECT * FROM 's3://my-public-bucket/data.parquet';

Private Buckets

To access private data, you must configure credentials in Settings → Remote Database.

1. Enter your Access Key ID and Secret Access Key.
2. Set the Region (e.g., us-east-1).
3. Set the Endpoint (leave blank for AWS, set for MinIO/R2).

Note: Credentials are stored locally in your browser.

Remote Databases (via Proxy)

To query PostgreSQL or MySQL, you cannot connect directly from the browser. You must use the PondPilot Proxy.

1. Deploy the Proxy or run it locally.
2. In PondPilot, go to Settings → Remote Database.
3. Add a Remote DuckDB connection pointing to your proxy URL.
4. Your Postgres/MySQL tables will appear in the Data Explorer as if they were local.