Hi friends of the Query Farm,

Grainlift now lets you query Cloudflare’s Durable Objects, D1, and Workers Analytics Engine from Haybarn, DuckDB, Python, or another ADBC application.

I’ve written up how it works, with a live demo you can try as you read about how it works.

Just to be predictable I put in a small coffee-shop database running in a Durable Object at Cloudflare. It's really remarkable how easy it is to use Durable Objects and D1. You can run SQL directly in the article, or open it in Cupola with five commented queries already in the editor (you don't even need to copy and paste or type rather). Browse the products, join orders to calculate revenue, or compare sales by city. No sign-in required or setup required. You should really think of the browser as a really nice way to have people access their data.

There are also examples for the Haybarn/DuckDB command line and Python. They all connect to the same database, which resets every four hours.

In the browser, Haybarn and the Grainlift ADBC driver run as WebAssembly. Grainlift uses vgi-rpc over HTTPS to reach Cloudflare and returns Apache Arrow results. You can also join those remote tables with data you bring into Haybarn or DuckDB.

The article walks through the architecture and how to connect your own Cloudflare data.

Read the post and try the demo

If you try it, hit reply and let me know if its delighted you today.

Best wishes,

Rusty