Getting Started

This page explains how to set up babyrite.

Requirements

Item	Minimum	Recommended (multi-guild)
CPU	1 vCPU (shared is fine)	1 vCPU
Memory	128MB	256MB or more
Disk	Tens of MB (config.toml only)	Same
Network	HTTPS outbound to the Discord Gateway (persistent WebSocket) and the Discord API / GitHub API	Same

• babyrite has no external database, and there are no plans to add a feature that would require one.
• When it connects to the Discord API over WebSocket, babyrite caches channel information in memory using moka.
  • The cache stores up to 500 entries each for a guild's channel list and individual channels.
  • It's configured with a 12h TTL and a 1h TTI.

These are implementation-derived estimates, not measured benchmarks

The figures above are estimated from the source code (cache size, the 1MB limit on GitHub raw fetches, etc.), not measured benchmarks. Adjust them to fit your deployment scale.

Supported OS / Architecture

• Verified to run on macOS and major Linux distributions, which are the recommended environments.
  • It also runs on Windows, but macOS or Linux is recommended (since v0.19.0).
• ARM64 environments are supported (since v0.16.0).

Trade-offs

babyrite's own processing is centered on regex matching and HTTP requests, so it rarely pins the CPU. That said, the following factors scale with the number of guilds and message volume, so running at the "minimum" spec across multiple guilds can lead to memory pressure or increased latency.

• moka cache memory usage: the more guilds and channels babyrite participates in, the more GuildChannel data the cache holds. The cache itself is capped at 500 entries, but the data size per entry depends on each guild's configuration.
• reqwest connection count: when Discord message links and GitHub permalinks are expanded concurrently (up to 3 per message each), outbound HTTP connections temporarily increase.
• GitHub raw fetch buffer: up to 1MB per file is temporarily loaded into memory, so momentary memory usage can spike when multiple GitHub permalinks are expanded at the same time.

The minimum spec is sufficient for small, single-guild deployments, but if you're running across multiple guilds continuously, it's recommended to leave some headroom at around the recommended spec.

Installation

babyrite is installed using Docker. You can pull the latest version with the following command.

docker pull ghcr.io/m1sk9/babyrite:v1

Using Docker Compose (recommended)

We recommend using Docker Compose to set up babyrite.

If you're using an orchestration tool like k8s or Docker Swarm, configure it to match that tool's own configuration files.

services:
  app:
    image: ghcr.io/m1sk9/babyrite:v1
    env_file:
      - .env
    restart: always

Configuration

You can customize babyrite's behavior with a dedicated configuration file. Write it in TOML format and point to it with the CONFIG_FILE_PATH environment variable. babyrite can also start with the default configuration if no configuration file is provided.

See the Configuration Reference for the full list of settings, their defaults, and logging output details.

Environment Variables

The environment variables babyrite uses are as follows. DISCORD_API_TOKEN is the only one required to start.

Key	Description
DISCORD_API_TOKEN	Discord API token
CONFIG_FILE_PATH	Path to the configuration file (recursive path)
RUST_LOG	Log level filter. When set, it overrides the configuration file's log.level.