Monitoring

The monitoring stack collects metrics from the device categories documented in this portal and stores the resulting time series in the shared metrics platform.

Architecture

graph LR
    A[Device Inventory] --> B[SNMP and Custom Exporters]
    C[Devices and Services] --> B
    B --> D[Prometheus]
    D --> E[Mimir]
    E --> F[Grafana Dashboards]
    E --> G[Grafana Alert Rules]
    G --> H[EMS / KeepHQ]

Prometheus uses inventory files for target discovery. Exporters translate SNMP, API, CLI, AXL/SOAP, PerfMon, or collectd data into Prometheus metrics. Prometheus remote-writes those samples to Mimir, and Grafana provides dashboards and metric alert evaluation.

Components

Component	Purpose
SNMP Exporter	Collects standard device and infrastructure metrics via SNMP polling
Custom Exporters	Collects device-specific API, AXL/SOAP, CLI, PerfMon, and collectd data
Mimir	Stores Prometheus-compatible time series for long-term query
Grafana	Provides dashboards, Explore, and metric-based alerts

Device Coverage

Collection methods differ by device. The Devices pages describe the expected exporter, log index, dashboard, and known gaps for each category.

Inventory presence is not live-health proof. When data is missing, check the path in this order: inventory → target discovery → exporter → Prometheus scrape → Mimir remote write → Grafana query.