Cilium Network Monitoring

Comprehensive Cilium network monitoring dashboard for Kubernetes clusters. Features instance filtering by Cilium instances (IP:port), endpoint state monitoring, BPF map usage, API performance metrics, and node connectivity status.

Cilium Network Monitoring Dashboard

This dashboard provides comprehensive monitoring for Cilium CNI in Kubernetes clusters.

Prerequisites

Cilium CNI installed and configured
Prometheus configured to scrape Cilium metrics
Cilium metrics enabled (default port 9962)

Configuration

Import this dashboard into your Grafana instance
Configure Prometheus as data source
Replace the following variables with your actual node IPs:
- NODE_IP_1 → Your first node IP
- NODE_IP_2 → Your second node IP
- NODE_IP_3 → Your third node IP
- NODE_IP_4 → Your fourth node IP

Metrics Required

The dashboard uses the following Cilium metrics:

cilium_agent_bootstrap_seconds
cilium_errors_warnings_total
cilium_agent_api_process_time_seconds_count
cilium_endpoint_state
cilium_bpf_map_capacity
cilium_kubernetes_events_received_total
cilium_bpf_map_ops_total
cilium_node_health_connectivity_status

Features

Bootstrap Monitoring: Track Cilium agent startup times
Error Tracking: Monitor errors and warnings across instances
API Performance: Track API response times and success rates
Endpoint States: Monitor endpoint readiness and regeneration
BPF Map Usage: Track memory usage of BPF maps
Node Connectivity: Monitor inter-node connectivity status

Dashboard Panels

Cilium Agent Bootstrap Time - Startup performance
Cilium Errors - Error count monitoring
Cilium Warnings - Warning count monitoring
Cilium API Success Rate - API performance
Cilium Errors & Warnings - Combined error tracking
Cilium Endpoint States - Endpoint status monitoring
BPF Map Memory Usage - Memory utilization
Cilium Agent API Process Time Rate - API latency
Kubernetes Events Received - Event processing
BPF Map Operations Rate - Map operation metrics 11-14. Node Health Connectivity - Per-node connectivity status