Ingress Nginx (Prometheus)¶

Display of Ingress performance Metrics, including average CPU usage of the Ingress Controller, average memory usage, total network requests/responses, number of times the Ingress Config has been loaded, result of the last Ingress Config load, and forwarding success rate of Ingress, etc.

Configuration¶

Prerequisites¶

• DataKit has been deployed; please refer to Kubernetes cluster <Install Datakit>

Installation and Deployment¶

Note: The example Ingress version is willdockerhub/ingress-nginx-controller:v1.0.0 (deployed under CentOS environment using kubeadmin). There may be differences in Metrics across different versions.

Metric Collection¶

• Obtain the yaml file for the deployed Ingress

wget https://raw.githubusercontent.com/kubernetes/ingress-nginx/controller-v1.1.1/deploy/static/provider/baremetal/deploy.yaml

• Edit deploy.yaml Set the service type to NodePort and expose port 10254.

spec:
  type: NodePort
......
    - name: prometheus
      port: 10254
      targetPort: prometheus

Find the part where kind: Deployment, and modify as follows:

kind: DaemonSet # Modified

---
hostNetwork: true # Added
dnsPolicy: ClusterFirstWithHostNet # Modified

• Enable Input To collect Ingress Metrics data in TrueWatch, you need to enable the prom plugin in DataKit. Specify the exporter's url in the prom plugin configuration. In a Kubernetes cluster, it is recommended to use annotations to add notes when collecting Ingress Controller Metrics. Open the deploy.yaml file used to deploy Ingress, find the DaemonSet section modified in the previous step, and add annotations.

annotations:
  datakit/prom.instances: |
    [[inputs.prom]]
      url = "http://$IP:10254/metrics"
      source = "prom-ingress"
      metric_types = ["counter", "gauge", "histogram"]
      # metric_name_filter = ["cpu"]
      # measurement_prefix = ""
      measurement_name = "prom_ingress"
      interval = "60s"
      tags_ignore = ["build","le","path","method","release","repository"]
      metric_name_filter = ["nginx_process_cpu_seconds_total","nginx_process_resident_memory_bytes","request_size","response_size","requests","success","config_last_reload_successful"]
      [[inputs.prom.measurements]]
        prefix = "nginx_ingress_controller_"
        name = "prom_ingress"
      [inputs.prom.tags]
      namespace = "$NAMESPACE"

• Parameter Description:

• url: Exporter urls, multiple urls separated by commas

• source: Collector alias
• metric_types: Metric types, optional values are counter, gauge, histogram, summary
• measurement_name: Measurement name
• interval: Collection frequency
• inputs.prom.measurements: Metrics with the specified prefix will be grouped into the specified name Measurement
• tags_ignore: Ignored tags

• metric_name_filter: Retained Metric names

• Deploy Ingress

kubectl apply -f deploy.yaml

Metrics¶

If inputs.prom.measurements is configured, the Metrics collected by TrueWatch need to have the prefix added to match the table.
For example, if the prefix nginx*ingress_controller is set, the Measurement is prom_ingress.

 [[inputs.prom.measurements]]
              prefix = "nginx_ingress_controller_"
              name = "prom_ingress"

The Metric nginx_ingress_controller_requests becomes the requests Metric under the prom_ingress Measurement in TrueWatch.