uptime/todo.md

arcline-uptime — Lightweight Uptime Monitor

Polls HTTP/TCP endpoints on a schedule, stores results in SQLite, sends alerts via webhook (Discord, Slack) and/or email. Single binary, no external services.

Stack

• Language: Go
• Storage: SQLite (via modernc.org/sqlite — pure Go, no CGO)
• Config: YAML
• Alerts: HTTP webhook (Discord/Slack compatible), SMTP email
• Optional UI: embedded web dashboard (net/http + Go templates)

Config format (uptime.yaml)

global:
  check_interval: 60      # seconds
  timeout: 10             # seconds per check
  alert_cooldown: 300     # seconds between repeat alerts for same monitor

alerts:
  - type: discord
    webhook_url: "https://discord.com/api/webhooks/..."
  - type: email
    smtp_host: mail.arclineit.com
    smtp_port: 587
    from: alerts@arclineit.com
    to: blake@arclineit.com

monitors:
  - name: "Main Website"
    type: http
    url: "https://arclineit.com"
    expected_status: 200
    contains: "[arcline]"        # optional string check in body

  - name: "Control Panel"
    type: http
    url: "https://cp.arclineit.com"
    expected_status: 200

  - name: "SSH"
    type: tcp
    host: "server1.arclineit.com"
    port: 22

  - name: "Mail Server"
    type: tcp
    host: "mail.arclineit.com"
    port: 587

Web dashboard

• / — current status of all monitors (live, auto-refresh)
• /history — response time graph (ASCII sparklines or simple SVG)
• /metrics — Prometheus-compatible text endpoint (optional)
• Protected by basic auth (config: dashboard.username / dashboard.password)

Alert format (Discord example)

[DOWN] Main Website
Expected 200, got 503
Checked at 2026-03-03 14:32:01 UTC
Response time: 8043ms (timeout)

Tasks

• Project scaffold
• YAML config parser
• HTTP monitor (status code, body contains, response time)
• TCP monitor (dial timeout)
• Scheduler (ticker per monitor, respect interval)
• SQLite schema (monitors, checks, alerts_sent)
• Result storage
• Discord webhook alerter
• SMTP email alerter
• Alert cooldown logic (don't spam on sustained outage)
• Recovery alert ("Main Website is back up, was down 12m 34s")
• Web dashboard — current status page
• Web dashboard — history / sparkline graph
• /metrics Prometheus endpoint
• Basic auth for dashboard
• systemd unit file example
• README with self-hosting guide
• Cross-compile Makefile