Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/abelperezr/nokia-bng-lab/llms.txt

Use this file to discover all available pages before exploring further.

​
Telemetry Overview

The Nokia BNG Lab includes a complete telemetry stack for real-time monitoring and visualization of network metrics using gNMI streaming telemetry.

​
Why Telemetry Matters

Telemetry provides:
  • Real-time visibility into network performance and health
  • Historical data for trend analysis and capacity planning
  • Automated monitoring without polling overhead
  • Granular metrics at 5-second intervals
  • Efficient streaming using gRPC and Protocol Buffers
Unlike traditional SNMP polling, gNMI streaming telemetry pushes data continuously from network devices, reducing CPU overhead and providing sub-second visibility into network operations.

​
Telemetry Architecture

The lab uses a modern observability stack: 
┌─────────────────────────────────────────────────────────────┐
│                   TELEMETRY DATA FLOW                       │
├─────────────────────────────────────────────────────────────┤
│                                                             │
│  ┌──────────┐   gNMI Stream    ┌──────────┐               │
│  │ BNG1/BNG2├─────────┐         │          │               │
│  │ (SROS)   │         │         │          │               │
│  └──────────┘         ├────────►│  gNMIc   │               │
│                       │         │ Collector│               │
│  ┌──────────┐         │         │          │               │
│  │ Switch/OLT├────────┘         └────┬─────┘               │
│  │ (SR Linux)│                       │                     │
│  └──────────┘                        │                     │
│                              Prometheus Export             │
│                              (Port 9273)                   │
│                                      │                     │
│                                      ▼                     │
│                              ┌──────────────┐              │
│                              │  Prometheus  │              │
│                              │  (TSDB)      │              │
│                              │  Port 9090   │              │
│                              └──────┬───────┘              │
│                                     │                      │
│                              PromQL Queries                │
│                                     │                      │
│                                     ▼                      │
│                              ┌──────────────┐              │
│                              │   Grafana    │              │
│                              │ Dashboards   │              │
│                              │  Port 3030   │              │
│                              └──────────────┘              │
│                                                             │
└─────────────────────────────────────────────────────────────┘

​
Components and Roles

​
gNMIc - Telemetry Collector

Role: Subscribe to gNMI streams from Nokia devices and convert data to Prometheus format Key Features:
  • Auto-discovers Containerlab nodes via Docker socket
  • Maintains persistent gRPC connections
  • Processes and transforms telemetry data
  • Exposes metrics in Prometheus format
Container: gnmic (10.77.1.12)
Port: 9273 (Prometheus exporter)
Config: configs/gnmic/config.yml
gNMIc automatically discovers all Nokia nodes in the lab and subscribes to their telemetry streams without manual configuration.

​
Prometheus - Time Series Database

Role: Store, query, and alert on time-series metrics data Key Features:
  • Scrapes metrics from gNMIc every 5 seconds
  • Stores historical data for trend analysis
  • Provides PromQL query language
  • High-performance time-series storage
Container: prometheus (10.77.1.13)
Port: 9090 (Web UI)
Config: configs/prometheus/prometheus.yml

​
Grafana - Visualization Platform

Role: Create dashboards and visualize telemetry data Key Features:
  • Pre-configured dashboards for SROS and SR Linux
  • Real-time metric visualization
  • Customizable panels and queries
  • Anonymous access enabled for easy viewing
Container: grafana (10.77.1.14)
Port: 3030 (Web UI)
Credentials: admin/admin
Config: configs/grafana/

​
Data Flow Explained

​
1. Telemetry Generation

Nokia devices continuously generate telemetry data for subscribed paths:
  • SROS devices (BNG1, BNG2): Interface stats, BNG sessions, system metrics
  • SR Linux devices (Switch, OLT, TX): Port stats, CPU/memory, BGP statistics

​
2. gNMI Streaming

Devices push data to gNMIc via gRPC: 
# Streaming mode with 5-second intervals
mode: stream
stream-mode: sample
sample-interval: 5s

​
3. Data Processing

gNMIc applies processors to:
  • Remove path prefixes (e.g., /state/)
  • Extract labels from paths
  • Convert state values (up=1, down=0)
  • Group metrics by interface

​
4. Prometheus Export

gNMIc exposes processed metrics at http://gnmic:9273/metrics

​
5. Metric Storage

Prometheus scrapes gNMIc every 5 seconds and stores time-series data

​
6. Visualization

Grafana queries Prometheus and renders dashboards with real-time data

​
Available Telemetry Data

​
SROS Metrics (BNG1, BNG2)

  • Interface operational state
  • Ingress/egress packets and bytes
  • Ethernet frame counters
  • Error counters
  • CPU utilization per slot
  • Memory usage and pools
  • Temperature sensors
  • Fan speeds
  • Local user database (LUDB) entries
  • IPoE session counts
  • PPPoE session counts
  • VPLS SAP statistics
  • BGP statistics and route counts
  • ISIS statistics
  • Route table entries (IPv4/IPv6)
  • Interface routing statistics

​
SR Linux Metrics (Switch, OLT, TX)

  • CPU utilization per control slot
  • Memory usage
  • Application management stats
  • Ethernet interface statistics
  • Subinterface counters
  • Operational state
  • Traffic rates
  • BGP neighbor statistics
  • Network instance route counts
  • LAG/LACP statistics

​
Getting Started with Telemetry

​
Step 1: Verify Stack is Running

# Check all telemetry containers are up
sudo docker ps | grep -E "gnmic|prometheus|grafana"

# Expected output: 3 running containers

​
Step 2: Access Grafana

Open your browser to: 
http://localhost:3030
No login required - anonymous access is enabled. To access admin features, use credentials: admin/admin

​
Step 3: Explore Dashboards

Navigate to Dashboards → Browse to see:
  • SROS Dashboard: BNG metrics, interface stats, system health
  • SR Linux Telemetry: Platform metrics, interface statistics

​
Step 4: Query Prometheus Directly

Access Prometheus UI at: 
http://localhost:9090
Try example queries: 
# CPU usage across all devices
system_cpu{}

# Interface operational state
port_oper_state

# BNG session count
subscriber_mgmt_local_user_db_ipoe_session_stats_current

​
Step 5: Verify gNMI Subscriptions

# Check gNMIc logs
sudo docker logs gnmic

# Look for successful subscriptions
# Example: "subscription initialized" for each device

​
Telemetry Configuration Files

All telemetry configuration is located in configs/: 
configs/
├── gnmic/
│   └── config.yml          # gNMIc subscriptions and processors
├── prometheus/
│   └── prometheus.yml      # Scrape configuration
└── grafana/
    ├── datasource.yml      # Prometheus datasource
    ├── dashboards.yml      # Dashboard provisioning
    └── dashboards/
        ├── SROS-Dashboard.json
        └── srlinux-telemetry-lite.json

​
Next Steps

Configure gNMIc

Learn about gNMI subscriptions and collectors

Query Prometheus

Write PromQL queries and explore metrics

Grafana Dashboards

Customize and create new dashboards

Available Metrics

Full catalog of Nokia SROS metrics

​
Troubleshooting

  1. Verify Prometheus is running: curl http://localhost:9090/-/healthy
  2. Check Prometheus targets: Navigate to http://localhost:9090/targets
  3. Ensure gNMIc is exposing metrics: curl http://localhost:9273/metrics
  1. Check gNMIc logs: sudo docker logs gnmic
  2. Verify devices are reachable: sudo docker exec gnmic ping bng1
  3. Check gNMI port is open on devices (57400)
  1. Check Prometheus config: sudo docker exec prometheus cat /etc/prometheus/prometheus.yml
  2. Verify network connectivity: sudo docker exec prometheus wget -O- http://gnmic:9273/metrics
  3. Review Prometheus logs: sudo docker logs prometheus
Prometheus stores metrics in memory. To reduce usage:
  • Decrease retention time (default: 15 days)
  • Reduce scrape interval in prometheus.yml
  • Limit subscription paths in gNMIc config

Build docs developers (and LLMs) love

Get started for freeTalk to us