Setting Up NPM Thresholding

This chapter describes performance monitoring using the NPM Threshold Management and Device Object Monitoring UIs.

After devices are discovered, you can monitor their availability and performance by defining NPM thresholds and mapping those thresholds to the appropriate device objects and measurements. This workflow typically starts with creating reusable threshold definitions (for example, device down, latency high, packet loss high, CPU high) in the NPM Threshold Management UI and then applying them to the discovered devices in the Device Object Monitoring UI to generate alerts when collected metrics breach the configured warning or critical conditions.

To open the NPM Threshold Management UI, from the main navigation menu, select Configuration, then Metrics, then NPM, and then NPM Thresholds.

To open the Device Object Monitoring UI, from the main navigation menu, select Configuration, then Metrics, then NPM, and then Device Objects.

Note:

To access these UIs, you need to deploy the Web UI and Web API microservices. See Setting Up Microservices for NPM.

About the NPM Threshold Management UI

Use the NPM Threshold Management UI to create and manage performance thresholds that can be reused across many devices. The UI provides search and a table of existing thresholds, with actions to create, enable or disable, edit, and delete thresholds (depending on your permissions).

When you create a threshold, you define the measurement and metric field it applies to (for example, fan and speed or ping and jitter), how it should be evaluated over time, and the alert conditions for Warning and Critical severities. After saving (and enabling) the threshold, it becomes available to map to device objects in the Device Object Monitoring UI.

Some examples are:

Device Down threshold with ping as measurement and availability as the field for which you will receive a critical alert when availability indicates down.
Latency High threshold with ping as measurement and latency as the field for which you will receive a warning and/or critical alert when latency exceeds a defined value.
Packet Loss High threshold with ping as measurement and packetLoss as the field for which you will receive a warning and/or critical alert when packet loss exceeds a defined value.
CPU High threshold with cpu as measurement and utilization as the field for which you will receive a warning and critical alert when CPU utilization exceeds defined levels.

Note:

Threshold values, operators, and supported fields depend on the measurement and metric field you select. Use consistent naming (for example, <Metric> High / <Metric> Low) so thresholds are easy to find when mapping.

About the Device Object Monitoring UI

Use the Device Object Monitoring UI to map thresholds to device objects. Mapping is what makes a threshold actionable for monitoring and alerting.

When you map thresholds, you select one device along with a measurement context and apply one or more thresholds that match that measurement and field.

To map a threshold:

In Device Object Monitoring UI, find the row for the device and measurement you want to monitor. For example, a device with Measurement ping.
Click the Action column for that row to open the Map Thresholds panel.
In Map Thresholds panel, review the Available list and select one or more thresholds.
Confirm the selection and click Map Selected Thresholds.

After mapping, the device object row reflects that thresholds are associated, and the system can generate alerts when conditions are breached.

Note:

Map only thresholds that are compatible with the selected Measurement (and applicable fields). If a threshold does not appear in the Available list, verify that its Measurement and Field align with the selected device object’s measurement context.

Troubleshooting Tips

If you cannot find a threshold while mapping:
- Confirm that the threshold is Enabled in Threshold Management.
- Confirm that the threshold Measurement matches the device object’s Measurement.
If a device object is missing from Device Object Monitoring UI:
- Confirm that the device was successfully discovered and the required discovery method (ping, SNMP, or API) is configured.
If alerts do not trigger as expected:
- Review the threshold Function, Time Range, Frequency, and the Warning or Critical conditions for correctness.
- Verify that the device is actively polling and returning metric data.

Title and Copyright Information

Network Performance Management Reporting Guide

G49449-01