Monitoring NGINX Ingress Controller with VTS metrics

Introduced in GitLab 9.5.

Note: NGINX Ingress version 0.16 and above have built-in Prometheus metrics, which are different than the VTS based metrics.

GitLab has support for automatically detecting and monitoring the Kubernetes NGINX Ingress controller. This is provided by leveraging the included VTS Prometheus metrics exporter in version 0.9.0 through 0.15.x.

Requirements

Prometheus integration must be active.

Metrics supported

Name Query
Throughput (req/sec) sum(rate(nginx_upstream_responses_total{upstream=~"%{kube_namespace}-%{ci_environment_slug}-.*"}[2m])) by (status_code)
Latency (ms) avg(nginx_upstream_response_msecs_avg{upstream=~"%{kube_namespace}-%{ci_environment_slug}-.*"})
HTTP Error Rate (%) sum(rate(nginx_upstream_responses_total{status_code="5xx", upstream=~"%{kube_namespace}-%{ci_environment_slug}-.*"}[2m])) / sum(rate(nginx_upstream_responses_total{upstream=~"%{kube_namespace}-%{ci_environment_slug}-.*"}[2m])) * 100

Configuring NGINX Ingress monitoring

If you have deployed NGINX Ingress using GitLab’s Kubernetes cluster integration, Prometheus automatically monitors it.

For other deployments, there is some configuration required depending on your installation:

  • NGINX Ingress should be version 0.9.0 or above, with metrics enabled.
  • NGINX Ingress should be annotated for Prometheus monitoring.
  • Prometheus should be configured to monitor annotated pods.

About managed NGINX Ingress deployments

NGINX Ingress is deployed into the gitlab-managed-apps namespace, using the official Helm chart. NGINX Ingress is externally reachable via the Load Balancer’s Endpoint.

NGINX is configured for Prometheus monitoring, by setting:

  • enable-vts-status: "true", to export Prometheus metrics.
  • prometheus.io/scrape: "true", to enable automatic discovery.
  • prometheus.io/port: "10254", to specify the metrics port.

When used in conjunction with the GitLab deployed Prometheus service, response metrics are automatically collected.

Manually setting up NGINX Ingress for Prometheus monitoring

Version 0.9.0 and above of NGINX Ingress has built-in support for exporting Prometheus metrics. To enable, a ConfigMap setting must be passed: enable-vts-status: "true". Once enabled, a Prometheus metrics endpoint begins running on port 10254.

Next, the Ingress needs to be annotated for Prometheus monitoring. Two new annotations need to be added:

  • prometheus.io/scrape: "true"
  • prometheus.io/port: "10254"

Managing these settings depends on how NGINX Ingress has been deployed. If you have deployed via the official Helm chart, metrics can be enabled with controller.stats.enabled along with the required annotations. Alternatively it is possible edit the NGINX Ingress YAML directly in the Kubernetes dashboard.

Specifying the Environment label

In order to isolate and only display relevant metrics for a given environment, GitLab needs a method to detect which labels are associated. To do this, GitLab searches for metrics with appropriate labels. In this case, the upstream label must be of the form <KUBE_NAMESPACE>-<CI_ENVIRONMENT_SLUG>-*.

If you have used Auto Deploy to deploy your app, this format is used automatically and metrics are detected with no action on your part.