Monitoring Kubernetes

Introduced in GitLab 9.0.

GitLab has support for automatically detecting and monitoring Kubernetes metrics.

Requirements

The Prometheus and Kubernetes integration services must be enabled.

Metrics supported

  • Average Memory Usage (MB):

      avg(sum(container_memory_usage_bytes{container_name!="POD",pod_name=~"^%{ci_environment_slug}-([^c].*|c([^a]|a([^n]|n([^a]|a([^r]|r[^y])))).*|)-(.*)",namespace="%{kube_namespace}"}) by (job)) without (job) / count(avg(container_memory_usage_bytes{container_name!="POD",pod_name=~"^%{ci_environment_slug}-([^c].*|c([^a]|a([^n]|n([^a]|a([^r]|r[^y])))).*|)-(.*)",namespace="%{kube_namespace}"}) without (job)) /1024/1024
    
  • Average CPU Utilization (%):

      avg(sum(rate(container_cpu_usage_seconds_total{container_name!="POD",pod_name=~"^%{ci_environment_slug}-([^c].*|c([^a]|a([^n]|n([^a]|a([^r]|r[^y])))).*|)-(.*)",namespace="%{kube_namespace}"}[15m])) by (job)) without (job) / count(sum(rate(container_cpu_usage_seconds_total{container_name!="POD",pod_name=~"^%{ci_environment_slug}-([^c].*|c([^a]|a([^n]|n([^a]|a([^r]|r[^y])))).*|)-(.*)",namespace="%{kube_namespace}"}[15m])) by (pod_name))
    

Configuring Prometheus to monitor for Kubernetes metrics

Prometheus needs to be deployed into the cluster and configured properly in order to gather Kubernetes metrics. GitLab supports two methods for doing so:

Specifying the Environment

In order to isolate and only display relevant CPU and Memory metrics for a given environment, GitLab needs a method to detect which containers it is running. Because these metrics are tracked at the container level, traditional Kubernetes labels are not available.

Instead, the Deployment or DaemonSet name should begin with CI_ENVIRONMENT_SLUG. It can be followed by a - and additional content if desired. For example, a deployment name of review-homepage-5620p5 would match the review/homepage environment.

Displaying Canary metrics

Introduced in GitLab 10.2.

GitLab also gathers Kubernetes metrics for canary deployments, allowing easy comparison between the current deployed version and the canary.

These metrics expect the Deployment or DaemonSet name to begin with $CI_ENVIRONMENT_SLUG-canary, to isolate the canary metrics.

Canary metrics supported

  • Average Memory Usage (MB)

      avg(sum(container_memory_usage_bytes{container_name!="POD",pod_name=~"^%{ci_environment_slug}-canary-(.*)",namespace="%{kube_namespace}"}) by (job)) without (job) / count(avg(container_memory_usage_bytes{container_name!="POD",pod_name=~"^%{ci_environment_slug}-canary-(.*)",namespace="%{kube_namespace}"}) without (job)) /1024/1024
    
  • Average CPU Utilization (%)

      avg(sum(rate(container_cpu_usage_seconds_total{container_name!="POD",pod_name=~"^%{ci_environment_slug}-canary-(.*)",namespace="%{kube_namespace}"}[15m])) by (job)) without (job) / count(sum(rate(container_cpu_usage_seconds_total{container_name!="POD",pod_name=~"^%{ci_environment_slug}-canary-(.*)",namespace="%{kube_namespace}"}[15m])) by (pod_name))