Skip to content

monitoring.md

Latest commit

 

History

History
189 lines (121 loc) · 6.52 KB

monitoring.md

File metadata and controls

189 lines (121 loc) · 6.52 KB
Raw
MyNet

Monitoring

Monitoring overview

MyNet pings every monitored device on a configurable schedule, tracks latency over time, and raises an alert when a device goes offline. WAN connections are monitored separately to track internet health.

Contents

How Monitoring Works

MyNet runs a background scheduler that pings all monitored devices at their configured intervals. Pings use ICMP (raw ping) via the icmplib library — no external tools required.

Each ping result records:

Field Description
Status up, down, or timeout
Latency Round-trip time in milliseconds (null on failure)
Timestamp When the ping was sent
IP Pinged The specific IP address that was pinged

Results are stored in the database and retained for 48 hours. The scheduler runs a cleanup job every 2 hours to remove old results.

Enabling Monitoring on a Device

  1. Edit the device (or use the monitoring toggle on the device detail page)
  2. Enable Monitoring
  3. Configure:
Setting Description Default
Interval How often to ping (seconds) 60
Monitor NICs Which NICs to ping All Ethernet NICs

When you save, the device is added to the monitoring scheduler immediately — no restart needed.

Which NICs are monitored: By default, all Ethernet (ETH) NICs with a valid IP address are monitored. You can select specific NICs in the device editor. Devices with only DHCP as their IP are skipped (no static IP to ping).

Monitoring Page

The Monitoring page (/monitoring) shows all monitored devices in a sortable list.

Each row displays:

Column Description
Device Name, type icon, and link to device detail
Status Current status badge: Online / Offline / Unknown
NICs Per-NIC status and latency sparkline
Uptime Percentage of successful pings in the last 24 hours
Avg Latency Mean round-trip time over the last 24 hours
Last Seen Timestamp of the last successful ping

Click a device row to expand it and see per-NIC latency history charts.

Sorting: Click column headers to sort by status, uptime, or latency.

Device Monitoring Detail

On any device detail page, the Monitoring tab shows:

  • Current status for each monitored NIC
  • 48-point sparkline chart showing latency over the last 24 hours
  • Uptime percentage
  • Average latency
  • Last seen timestamp
  • Total pings sent / received

The Ping Now button (Quick Actions) triggers an immediate on-demand ping regardless of the scheduled interval.

WAN Monitoring

WAN connections are monitored separately from LAN devices. WAN monitoring is configured on each WAN port of a router or gateway device.

How it works:

  • Each WAN port can have a ping target IP (default: 1.1.1.1)
  • MyNet pings this target to test internet connectivity
  • Results are tracked per WAN connection

WAN Status appears on the Dashboard and in the monitoring summary:

  • Total WAN connections configured
  • How many are currently reachable
  • Per-connection status

When a WAN connection goes offline, a warning event is raised. When it recovers, a recovery event is logged.

Real-Time Updates

Monitoring status updates are pushed to the browser via WebSocket. You do not need to refresh the page — status badges and sparklines update automatically.

The WebSocket connection requires authentication. If you are logged in, it connects automatically.

On-Demand Ping

Click Ping Now on any device detail page to immediately send a ping, outside of the normal schedule. The result appears in the monitoring panel within a few seconds.

This is useful for:

  • Testing whether a device just came online
  • Verifying a network change took effect
  • Checking latency after moving a device

Requires Editor role.

Alerts & Events

MyNet raises events automatically based on monitoring results:

Event Trigger
Device Offline A device fails N consecutive pings (default: 3)
Device Recovered A previously-offline device responds again
WAN Offline A WAN ping target becomes unreachable
WAN Recovered A WAN connection is restored

These events appear in the Events log and are included in the active alert count shown in the header.

The failure threshold (N) is configured via MONITORING_FAILURE_THRESHOLD in .env (default: 3).

Performance Tuning

Concurrent pings

By default, up to 20 pings run concurrently. For faster monitoring of many devices, increase this in .env:

MONITORING_MAX_CONCURRENT_PINGS=50

On a Raspberry Pi 3B+, a safe ceiling is about 150 concurrent pings. Beyond that, the 60-second scheduler tick may take longer than the tick interval.

Monitoring interval

The scheduler ticks every 60 seconds. Each device is pinged no more frequently than its configured interval. The default interval for new devices is 60 seconds.

For large installations (300+ devices), consider:

  • Setting intervals to 120s or longer
  • Reducing result retention from 48h to 24h (change the hours=48 value in cleanup_old_results() in services/monitoring_scheduler.py)
  • Moving the database to a faster medium (USB SSD instead of SD card on Pi)

Raspberry Pi SD card wear

At 85 devices with 60-second intervals, MyNet writes approximately 70–100 MB/day to SQLite. Recommendations:

  • Move the database to a dedicated USB drive via the built-in Storage feature (Settings → Storage). MyNet handles the migration, keeps hourly SD snapshots for recovery, and surfaces USB health in the UI — no manual .env edits or file moves.
  • Use a high-endurance SD card (Samsung Pro Endurance, SanDisk High Endurance)
  • Increase TICK_SECS from 60 to 120 in services/monitoring_scheduler.py if your monitoring needs tolerate a slower tick

← Networks · Events →