etcd-proxy-server

Proxy to collect etcd metrics using Prometheus over HTTPS in a Kubernetes cluster

• Automatically deployed on master nodes (daemonSet.yaml);
• Using certificates located in /etc/kubernetes/pki/etcd, configures the HTTPS client;
• Receives requests for receiving metrics and, on behalf of the configured client, refers to etcd;
• Returns received metrics;
• Allows only GET requests for the handler /metrics;

Usage

Kubernetes

Just do it and forget it:

git clone git@github.com:eikoshelev/etcd-proxy-server.git && cd etcd-proxy-server

kubectl apply -f kubernetes -n <namespace>

Docker container

docker run -d --name etcd-proxy-server eikoshelev/etcd-proxy-server:v2.0.0

Description

./etcd-proxy-server -h

Usage of ./etcd-proxy-server:
  -caFile string
    	A PEM eoncoded CA's certificate file (default "/etc/kubernetes/pki/etcd/ca.crt")
  -certFile string
    	A PEM eoncoded certificate file (default "/etc/kubernetes/pki/etcd/healthcheck-client.crt")
  -clientTimeout duration
    	Timeout for client (default 10s)
  -hostIP string
    	Host machine IP
  -keyFile string
    	A PEM encoded private key file (default "/etc/kubernetes/pki/etcd/healthcheck-client.key")
  -metricsRoute string
    	Route for collecting metrics (default "/metrics")
  -serverPort string
    	Server port (default ":8888")
  -serverRTimeout duration
    	ReadTimeout for server (default 10s)
  -serverWTimeout duration
    	WriteTimeout for server (default 10s)

hostIP reads the environment variable of the same name by default, which is set depending on the node on which it is deployed under, for more details see daemonSet.yaml