To start getting metrics from a set of nodes, execute this on every node
curl -o- https://raw.githubusercontent.com/JayjeetAtGithub/prometheus-on-baremetal/master/exporter.sh | bash
Write the host names in the hosts file and generate the prometheus config file in /tmp by doing,
git clone https://github.com/JayjeetAtGithub/prometheus-on-baremetal
cd prometheus-on-baremetal/
python3 get_config.py
To start prometheus on the admin node,
curl -o- https://raw.githubusercontent.com/JayjeetAtGithub/prometheus-on-baremetal/master/prometheus.sh | bash
To start Grafana server,
curl -o- https://get.docker.com | bash
docker run -d -p 3000:3000 grafana/grafana
After starting the Grafana server, add a prometheus data source named skyhook and then import the provided grafana.json to get a pre-defined dashboard to monitor CPU, Storage and Network usage across the cluster.