How I choose between D1, KV, and R2 on Cloudflare Workers for indie SaaS

After building several small bootstrapped SaaS tools on Cloudflare Workers, I put together a clear decision framework for its three core edge storage services:

• D1 for relational structured business data

• KV for global low-latency caching and sessions (with caveats around eventual consistency)

• R2 for file uploads and binary assets with zero egress cost

I also covered common pitfalls like misusing KV for user persistent data, ignoring TTL rules, and mixing up blob storage locations.

Would love to hear how other indie founders architect their data layers on the Cloudflare edge runtime.