metrics-server-exporter
Metrics Server Exporter
The metrics server exporter is solution to help you monitor RAM/CPU easily.
You can monitor RAM/CPU of the windows container via the metrics server exporter
Badge
Helm Chart:
Docker:
Prequirement
To the metrics-server-exporter, you need prepare something:
- Install Kubernetes Metrics Server
Install metrics-server-exporter
We will install the metrics-server-exporter through Helm Chart.
helm repo add openvscode-server-helm https://mrnim94.github.io/metrics-server-exporter
helm search repo metrics-server-exporter
NAME CHART VERSION APP VERSION DESCRIPTION
metrics-server-exporter/metrics-server-exporter 0.1.2 0.0.1 A Helm chart for Kubernetes
Importance
Default metrics-server-exporter will collect the usage resources of ALL namespace
Besides, you can specify many namespace via environment and metrics-server-exporter will collect the metrics follow
your mean
VALUES FILES
#values.yaml
envVars:
LOOK_NAMESPACES: <namespace1>,<namespace2>
Metrics
http://localhost:1994
| Name | Description | Type | Labels |
|---|---|---|---|
metrics_server_pod_cpu_usage |
Metrics server pod cpu utilization (m or Millicore) | gauge | namespace, kind_owner, name_owner, pod, container, os |
metrics_server_pod_memory_usage |
Metrics server pod memory utilization (Mi or Mebibyte) | gauge | namespace, kind_owner, name_owner, pod, container, os |
metrics_server_hpa_utilization |
Current Average Utilization Percentage of each metric that is created by HPA (%) | gauge | namespace, metric_name, metric_type, hpa_owner, scale_target_ref_kind, scale_target_ref_name |
Grafana Daskboard
- Get Grafana Json
Refer to:
Default to enable Auto Discovery Metrics
We are configured prometheus.io/scrape: "true" in Service template.
Maybe your Promethues will be automatic to get metrics and you don’t need to configure anything.
Create Helm Package
helm package ./helm-chart/metrics-server-exporter --destination ./helm-chart/
helm repo index . --url https://antiantiops.github.io/metrics-server-exporter