Skip to content
Fleet

Alerts & Monitoring

fleet-bot includes four monitoring systems that run continuously and send alerts through your configured adapter.

Alert Monitor

Polls fleet health every 2 minutes and alerts on state transitions.

  • Tracks per-app health state (healthy, degraded, down)
  • Only alerts when state changes (no repeated alerts for the same issue)
  • Auto-restart: optionally restarts apps that go down (with 10-minute cooldown between restarts)
  • Mute non-critical: when enabled, only alerts on down transitions (ignores degraded)

Commands

/alerts              — show current settings + toggle buttons
/alerts on|off       — enable/disable monitoring
/alerts autorestart  — toggle auto-restart
/alerts mute         — toggle mute non-critical

Ping Monitor

HTTP-pings all app domains every 3 minutes.

  • Checks each app’s configured domains with a 10-second timeout
  • Alerts on failures (connection error, timeout, 5xx status)
  • Alerts on slow responses (> 3 seconds)
  • Tracks per-domain status and latency

Commands

/ping          — show all ping results
/ping on|off   — enable/disable

Uptime Tracker

Polls app health every 2 minutes and persists uptime data to /etc/fleet/uptime.json.

  • Records upticks/downticks per app
  • Calculates uptime percentage over time
  • Data survives bot restarts (persisted to disk)

Commands

/uptime        — show uptime percentages for all apps

Daily Digest

Sends a summary message at 08:00 UTC daily.

  • App health overview (healthy/degraded/down counts)
  • Uptime percentages for the past 24 hours
  • Ping latency summary
  • Alert count since last digest

Commands

/digest         — show digest settings
/digest on|off  — enable/disable daily digest

Scheduled deployments

/deploy_at 14:30 myapp

Schedules a deployment at the specified time. The digest manager handles the scheduling.

Alert destination

Alerts are sent to the first entry in alertChatIds (Telegram) or allowedNumbers (BlueBubbles). Configure this in /etc/fleet/bot.json.