Access Logging

Rustinx can log every request to a file. Logging is disabled by default.

Enable Logging

[logging]
enabled = true
path = "/var/log/rustinx/access.log"
timezone = "+01:00"

Settings

Timezone

Set the timezone offset for log timestamps. Use UTC offset format:

[logging]
enabled = true
timezone = "+01:00"  # Central European Time

If timezone is not set, timestamps are in UTC.

Log Format

Each line contains:

[hh:mmAM/PM | dd/mm] client_ip METHOD /path status response_time

Example output with timezone = "+01:00" (CET):

[03:37PM | 23/05] 192.168.1.1 GET /index.html 200 0.3ms
[03:37PM | 23/05] 10.0.0.5 GET /assets/style.css 200 0.1ms
[03:38PM | 23/05] 172.16.0.1 HEAD /api/health 200 0.0ms
[03:38PM | 23/05] 192.168.1.1 GET /missing 404 0.0ms

Client IPs come from X-Forwarded-For when behind_proxy = true, otherwise from the TCP connection.

Docker Setup

When running in Docker, mount a volume for the log directory:

services:
  web:
    image: ghcr.io/shadowarcanist/rustinx:v1.0
    volumes:
      - ./rustinx.toml:/etc/rustinx/rustinx.toml:ro
      - ./dist:/static:ro
      - ./logs:/var/log/rustinx    # writable log volume

The container filesystem is read-only, so you must mount the log directory separately.

How It Works Internally

Logging uses an async channel to avoid blocking request handling:

• Each request sends a log line to a bounded channel (4096 entries)
• A dedicated writer task drains the channel and writes to disk
• If the channel fills up during traffic spikes, excess log entries are dropped
• Drop events are reported to stderr with a count
• On shutdown, the writer writes any remaining entries to disk

This means logging never slows down request serving, even on slow disks.

Log Rotation

Rustinx does not rotate logs internally. Use your container runtime's log driver or an external tool:

{
  "log-driver": "json-file",
  "log-opts": {
    "max-size": "10m",
    "max-file": "3"
  }
}

Or use logrotate with the copytruncate directive for the mounted log file.