R2 + Worker Runtime

Core idea

The Worker does not fetch the whole object. It asks R2 only for the byte ranges it actually needs.

Why is that interesting?

• the amount of transferred data becomes smaller
• known query paths can be optimized very precisely
• runtime logic can stay at the edge close to the request

What does the Worker do?

In practice the Worker acts as a query planner and response assembler:

1. interprets the request
2. selects the required range reads
3. assembles the response from them

End-to-end path

HTTP request
  -> worker.js
  -> queryJdbinTop(...)
  -> R2 range reads
  -> DataView decode
  -> JSON response with rows + metrics

Why Cloudflare specifically?

Cloudflare provides the primitives this architecture needs:

• Worker
• R2
• byte-range reads
• edge execution

Important boundary

The Worker is not just an API proxy. It acts as planner, range reader, parser, and response builder.