Prometheus HTTP Exporter

Description

Prometheus exporter for gathering NGINX ingress endpoints by regex and check their status code using http path.

Metrics can be served or/and can be pushed to Prometheus Gateway.

Metics name: http_exporter_probe_http_status_code

Configuration

Example of config in ./chart/values.yaml

# Prometheus HTTP Exporter configuration file.
config:
  publish_port: 8000 # [Optional] Serving port, here you can see your metrics, default 8000
  check_delay: 5 # [Optional] How often we should check enpoints, in seconds, default 5
  push_gateway: # [Optional] Prometheus Pushgateway configuration
    address: prometheus-pushgateway:9091 # Address of Prometheus Pushgateway
    job: prometheus_http_exporter # [Optional] Metric label
  targets: # List of targets to gather and check
  - pattern: ^api.* # Regular expresion which will find your Ingress Host
    path: /swagger-ui.html # Path to check in each finded Ingress Host
    follow_redirects: true # [Optional] Should we follow redirects, default True
    insecure_skip_verify: true # [Optional] Should we check secure connection or not, default False
    timeout: 5 # [Optional] Fail if can't connect to URL after N seconds, default 5

Usage

Reguirenments

• Kubernetes Cluster 1.16+
• Helm 3+
• [Optional] Prometheus Gateway in K8S cluster
• [Optional] Grafana

Steps

1. Clone repository: git clone [email protected]:yevhen-kalyna/prometheus_http_exporter.git
2. Enter in it: cd prometheus_http_exporter
3. Change chart values: nano ./chart/values.yaml
4. Install Helm chart in your cluster: helm install prometheus-http-exporter -n <YOUR_NAMESPACE> ./chart
   1. [Optional] Or if installed, upgrade chart with: helm upgrade prometheus-http-exporter -n monitoring ./chart
5. [Optional] Check your new metrics in your Prometheus via name: http_exporter_probe_http_status_code
6. [Optional] If your Prometheus connected as datasourse to Grafana, you can query them: e.x. count by (url) (http_exporter_probe_http_status_code{url=~ ".*api[.].*"} == 200)