Monitoring for Nginx-Ingress on Kubernetes: only 3 metrics available ?

[m.przy...@biotcloud.com]

Hello,

I've started (or i would like to .. ) to monitor metrics/stats for my Nginx-Ingress controller on Kubernetes. I've enabled stats/metrics in configs, in config map also VTS is set to true - unfortunately in Prometheus i see only 3 metrics type --> nginx_connecticons, nginx_connecticons_total, nginx_requests_total <-- i believe something is wrong as there should be in my opinion much more, am i correct ? If so, what should i correct on my side as i'm out of ideas. On sample grafana dashboards i've seen there was a much more different metrics to choose.

I.e. when I curl into http://localhost:18080/nginx_status (after port forwarding on pod) i see only this in output:

Active connections: 7 
server accepts handled requests 
16792 16792 8067 
Reading: 0 Writing: 2 Waiting: 5 

My configuration for targets in Prometheus looks like following:

  - job_name: nginx-ingress-staging-stats
    scrape_interval: 20s
    scrape_timeout:  15s
    metrics_path: "/metrics"
    static_configs:
    - targets:
      - nginx-ingress-staging-controller-metrics.nginx-ingress:9913

IMO config for Nginx-Ingress chart should be ok .. 

  stats:
    enabled: true

    service:
      annotations: {}
      clusterIP: ""
      externalIPs: []

      loadBalancerIP: ""
      loadBalancerSourceRanges: []
      servicePort: 18080
      type: ClusterIP

  metrics:
    enabled: true

    service:
      annotations:
        prometheus.io/scrape: "true"
        prometheus.io/port: "10254"

      clusterIP: ""
      externalIPs: []

      loadBalancerIP: ""
      loadBalancerSourceRanges: []
      servicePort: 9913
      type: ClusterIP

[m.przy...@biotcloud.com]

All fixed, since 0.16.1 version of Nginx VTS metrics were removed. Prometheus metrics available after updating to 0.17.1.