Rewrite proxies and upstream redirect loops
A rewrite is a proxy, not a redirect, so it passes upstream 3xx responses straight through — why pointing one at an origin that canonicalizes creates an infinite loop, and how to fix it
Knowledge documents in Network.
A rewrite is a proxy, not a redirect, so it passes upstream 3xx responses straight through — why pointing one at an origin that canonicalizes creates an infinite loop, and how to fix it
The negotiation cache that uses ETag and If-None-Match to serve unchanged responses as 304, driving download bytes to zero. Covers the limit of saving only bandwidth (not server compute) and the cheap validator
Cache key normalization when storing temporary signed URLs: strip auth queries only, keep semantic queries
How to block SSRF with a scheme/host allowlist in a proxy where the server fetches a user-supplied URL
How the client uploads a binary directly to object storage using a one-time URL signed by the backend, and the step-by-step contract