Concepts

A few words, always the same, throughout the panel and this guide.

Domain

The DNS name visitors use to reach the files (media.company.com). A domain is verified when (1) DNS resolves it and (2) an HTTP request going through that name really reaches this node. Only verified domains can be used in routes. → Domains and DNS

Bucket

An S3 storage with its credentials and the name of the bucket to read from. In the panel a "bucket" gathers endpoint, region, addressing, keys and bucket name; keys are never shown again after saving. → S3 buckets

Route

The rule that links a domain + a path prefix to a bucket + a folder. With the route media.company.com + /docs/ → bucket documents, folder public/, the request media.company.com/docs/prices.pdf reads documents/public/prices.pdf. → Routes

Cache policy

How long a copy is fresh (ttl), how long a missing file is remembered (ttl_not_found), how long expired copies are served if the storage is down (serve_stale_on_error) and which query parameters matter (query_keys). The panel creates a standard policy: 1 hour, 60 seconds, 24 hours, no query parameters. → Cache

Configuration and version

The gateway works on a config.yaml file with a version that increases at every change. The panel regenerates it entirely from the data it manages (domains, buckets, routes) and applies it immediately; if you write the file by hand the panel recognizes it and leaves it alone (hand_managed). → config.yaml

Node

An OtterRoute instance with its own cache and state folders. Each node has an identity (node-id) with which it proves, during verification, that it is the one answering on a domain.

Users, roles, scopes

Who accesses the panel and what they can do: a role (Administrator, Operator, Read-only, Custom) is a set of scopes, that is, of allowed actions. → Users, permissions and 2FA