System Monitor

EdgeFlow includes a comprehensive monitoring system that tracks hardware resources, flow execution metrics, and system health in real-time. Monitor CPU, memory, disk, temperature, and network usage from the built-in dashboard.

Monitor Overview

The System Monitor dashboard provides an at-a-glance view of your device's health and performance. Circular gauges show real-time utilization with color-coded thresholds that shift from green to yellow to red as resources approach their limits.

System Metrics

EdgeFlow collects system metrics from the Linux /proc and /sys filesystems, providing detailed insight into hardware utilization without external dependencies.

CPU Monitoring

CPU metrics are gathered from Go's runtime package and /proc/loadavg, polled every 3 seconds by default.

Memory Usage

Memory information is read from /proc/meminfo and displayed as a segmented breakdown bar. The monitor tracks total, used, available, and swap partitions.

Disk Usage

Disk utilization is reported per-partition. EdgeFlow monitors the root filesystem and any mounted data partitions.

Temperature Monitoring

On Raspberry Pi and other supported boards, EdgeFlow reads the CPU temperature from /sys/class/thermal/thermal_zone0/temp. The value is displayed as a gauge with color-coded zones indicating thermal state.

Network

Network statistics are read from /proc/net/dev and polled at the same interval as other system metrics. Speed is calculated as the byte delta between consecutive polls.

Monitoring API Endpoints

All monitoring data is accessible through REST API endpoints. These are the same endpoints used by the built-in dashboard and can be consumed by external tools.

Example Response

GET /api/v1/system/info

Response:
{
  "hostname": "edgeflow",
  "os": "linux",
  "arch": "arm64",
  "board": "Raspberry Pi 4 Model B",
  "uptime": "12d 5h 23m",
  "temperature": 52.1,
  "load_avg": [0.45, 0.38, 0.32],
  "memory": {
    "total": 1073741824,
    "used": 163577856
  }
}

Resource Limits & Auto-Management

EdgeFlow defines soft and hard resource thresholds in its ResourceLimits configuration. When thresholds are crossed, the system takes progressive action from logging warnings to automatically disabling non-essential modules.

When memory usage exceeds the hard limit, EdgeFlow automatically disables non-essential modules to reclaim resources. The CanLoadModule() function checks available memory before allowing new modules to be loaded, preventing out-of-memory situations on constrained devices.

Health Check System

The health check system runs periodic checks against critical subsystems and reports an aggregate status. Each check returns one of three severity levels:

Built-in Health Checks

The health endpoint is available at GET /api/v1/health and returns the aggregate status along with individual check results. Periodic checks run automatically on a configurable interval (default: 30 seconds).

GET /api/v1/health

{
  "status": "healthy",
  "checks": {
    "database": { "status": "healthy", "latency_ms": 2 },
    "disk": { "status": "healthy", "usage_percent": 45 },
    "memory": { "status": "healthy", "usage_percent": 35 },
    "goroutines": { "status": "healthy", "count": 42 }
  }
}

Flow & Execution Metrics

EdgeFlow exposes Prometheus-format metrics for flow execution, node activity, and system performance. These metrics can be scraped by Prometheus and visualized in Grafana.

Prometheus & Grafana Integration

Enabling Prometheus Metrics

Enable Prometheus metrics export in Settings > Metrics. Once enabled, EdgeFlow exposes a /metrics endpoint on the configured Prometheus port (default: 9090) that returns all metrics in Prometheus text format.

Add EdgeFlow as a scrape target in your prometheus.yml configuration:

scrape_configs:
  - job_name: 'edgeflow'
    static_configs:
      - targets: ['raspberrypi:9090']
    scrape_interval: 15s

Grafana Dashboard

To visualize EdgeFlow metrics in Grafana:

1. Open Grafana and navigate to Configuration > Data Sources
2. Click Add data source and select Prometheus
3. Set the URL to your Prometheus server (e.g., http://localhost:9090)
4. Click Save & Test to verify the connection
5. Import the EdgeFlow dashboard template from Dashboards > Import
6. Select the Prometheus data source and click Import

Log Viewer

The built-in Log Viewer streams real-time log entries via WebSocket, supporting log, flow_status, and node_status event types. It retains the last 500 entries in the browser for scrollback.

Log Viewer Features

• Level Filter - Filter by ALL, DEBUG, INFO, WARN, or ERROR
• Search - Full-text search across log messages
• Pause / Resume - Freeze the stream to inspect entries
• Auto-scroll - Automatically scrolls to newest entries when active
• Retention - Keeps the last 500 entries in the browser

Log Sources

Level Indicators