Laravel Server Monitor Laravel Package

Getting Started

Start by installing the package via Composer and publishing its migrations and config file. Run php artisan vendor:publish --provider="Spatie\ServerMonitor\ServerMonitorServiceProvider" twice (once for migrations, once for config), then php artisan migrate. Next, add your servers to the hosts table (or seed them), and configure SSH access (SSH key-based auth is strongly recommended). Register the server-monitor:run-checks command in your console kernel to run every minute. The first thing you’ll likely want to do is enable built-in checks like diskspace, mysql, and elasticsearch in config/server-monitor.php.

Implementation Patterns

Use Eloquent models (Host, Check) directly for programmatic management: e.g., Host::create(['name' => 'prod-web-01', 'hostname' => '10.0.0.5']). Create custom checks by extending CheckDefinition and implementing run() and parseOutput() — ideal for app-specific health signals (e.g., checking Laravel Horizons status, Redis queue depths, or custom health endpoints via localhost curl). Schedule recurring checks on a per-host basis by setting the next_run_in_minutes attribute on the Host model. Leverage events (e.g., CheckFailed, CheckRestored) to integrate with internal tools (e.g., PagerDuty, Opsgenie, or custom alerting dashboards) via listeners. Use the Spatie\ServerMonitor\CheckDefinition type-hinted interfaces and traits (HasCustomAttribute, IsRunnable) to safely inject config or environment-specific parameters.

Gotchas and Tips

SSH key permissions are critical — ensure the web server user (e.g., www-data) has read access to the private key (not world-readable for security). Set concurrent_ssh_connections low (e.g., 2–3) on shared dev/staging infrastructure to avoid lockouts; raise it on production only if needed and with proper SSH hardening. Use environment variables for sensitive values (SERVER_MONITOR_SLACK_WEBHOOK_URL, SSH keys) and never commit secrets. Tip: Wrap ssh in a wrapper script with set -e and custom timeouts to catch hanging connections (the package times out by default at ~15s). Custom check tip: Log verbose output via Laravel\Logs\Logger or standard output (which is captured and stored) to aid debugging. Remember: notifications only fire on state transitions (OK → warning/fail, fail → warning, etc.) and are throttled by throttle_failing_notifications_for_minutes. Extension point: use the process_manipulator config to inject retry logic, environment cleanup, or anonymization of sensitive hostnames before SSH execution.