Device Monitor
Device Monitor is the registry and health system for everything on your network that isn't an edge node: switches, firewalls, PLCs, HMIs, protocol gateways, and sensors. It registers them, polls them, and tells you when one stops answering.
Edge nodes are handled by Ops. Device Monitor covers the equipment around them.
The health panel
Devices live inside your site hierarchy, so you reach them by selecting a site and then a building. Four counters summarize what you find:
| Counter | Meaning |
|---|---|
| Healthy | Responding to health checks normally |
| Warning | Responding, but with elevated error rates or latency |
| Critical | Unreachable, or reporting a fault |
| Unknown | Registered but never successfully polled |
Unknown is worth treating differently from Critical. Critical means something that was working has stopped. Unknown almost always means the device was registered with the wrong address or the wrong SNMP credentials and has never worked at all.
Registering devices
You can register devices one at a time or by bulk import. Each one needs an address, a type, and, if you want more than reachability, SNMP credentials.
| Type | Typical use |
|---|---|
| Switch | Managed Ethernet switches |
| Firewall | Network security appliances |
| PLC | Programmable logic controllers |
| HMI | Human-machine interfaces |
| Gateway | Protocol converters |
| Sensor | Environmental sensors |
| Custom | Any other IP-connected device |
The type drives which panels you get. Registering a switch as Custom still monitors it, but you lose the port grid.
What polling gives you
With reachability only, you get up or down. With SNMP configured, you also get:
- Port grid, a visual map of switch port status and utilization
- Interface status, per-port link state, speed, errors, and traffic counters
- Sparklines, recent trend for each metric alongside its current value
Health checks run on a periodic schedule, and you can also trigger one on demand when you want an answer now rather than at the next interval.
Industrial protocol metrics
Beyond SNMP, Device Monitor can collect from equipment over the protocol it already speaks. Sixteen are supported:
| Protocol | Transport | Typical equipment |
|---|---|---|
| OPC UA | TCP 4840 | PLCs, SCADA, industrial gateways |
| Modbus TCP | TCP 502 | Drives, meters, sensors |
| BACnet/IP | UDP 47808 | Building automation, HVAC |
| EtherNet/IP | TCP 44818 | Allen-Bradley and Rockwell controllers |
| PROFINET | Ethernet | Siemens PLCs, distributed I/O |
| DNP3 | TCP 20000 | Power grid, water systems |
| MQTT | TCP 1883 / 8883 | IoT sensors, edge gateways |
| S7comm | TCP 102 | Siemens S7 PLCs |
| FINS | TCP/UDP 9600 | Omron PLCs |
| MC Protocol | TCP 5000 | Mitsubishi PLCs |
| ADS | TCP 48898 | Beckhoff TwinCAT |
| CIP | TCP 44818 | Common Industrial Protocol devices |
| HART-IP | TCP 5094 | HART field devices |
| IEC 61850 | TCP 102 | Power substations |
| IEC 60870-5-104 | TCP 2404 | Telecontrol and power SCADA |
| KNX/IP | UDP 3671 | Building automation |
Alerting
Device health transitions raise alerts through the normal pipeline, so a device going Critical reaches you the same way a node going offline does. See Alerts & Notifications.
Roles
| Role | What they can do |
|---|---|
| Global Command | Full management across tenants: register, edit, delete, configure SNMP and credentials |
| Admin | View health and metrics for their tenant, run on-demand checks |
| Engineer | View health and metrics, read-only |
| Operator | View device cards for their sites |
Next steps
- Network Devices: managing configuration on network equipment
- Alerts & Notifications: how device alerts are delivered
- Connectivity: connecting devices in the first place