Track NGINX Ingress Controller health: request volume, active connections, config reload status, resource usage, and per-ingress success rates.
This dashboard requires nginx-ingress-controller metrics flowing into SigNoz. Follow the Send NGINX Ingress Controller Metrics to SigNoz guide to configure the OpenTelemetry Collector to scrape the controller's Prometheus endpoint.
Dashboards → + New dashboard → Import JSON
What This Dashboard Monitors
- Request volume: total requests per second at the controller and per-ingress level
- Connections: active controller connections over time
- Success rate: fraction of non-4xx/5xx responses at controller and per-ingress level
- Config reloads: reload frequency and last-reload failure status
- Resource usage: average CPU and RSS memory per controller pod
- Network I/O: bytes sent and received
- Response times: p50/p95/p99 latency per ingress resource
Metrics Included
Controller Health
- Controller Request Volume: Shows requests per second across all ingresses on this controller over time
- Controller Connections: Shows active controller connections over time
- Controller Success Rate: Shows fraction of non-4xx/5xx responses (0–1 scale) over time
- Config Reloads: Shows rate of nginx config reloads over time
- Last Config Failed: Shows whether the most recent config reload failed over time
Per-Ingress
- Ingress Request Volume: Shows requests per second for the selected ingress resource over time
- Ingress Success Rate: Shows non-4xx/5xx response fraction for the selected ingress over time
- Ingress Percentile Response Times: Shows p50, p95, and p99 end-to-end request duration for the selected ingress over time
Resource Usage
- Average CPU Usage: Shows mean CPU consumption per controller pod over time
- Average Memory Usage: Shows mean RSS memory in bytes per controller pod over time
- Network I/O pressure: Shows bytes sent and received over time
Dashboard Variables
Variables cascade in order — each filter narrows the next:
- namespace: Kubernetes namespace where the controller runs
- controller_class: Ingress class (e.g.
k8s.io/ingress-nginx) - controller: Specific controller pod name
- ingress: Ingress resource name to drill into per-ingress panels
