Grafana is used to receive metric information from servers in production such as load, memory usage, networking and more. It is an essential tool that is used for graphing the data and then using it to evaluate outages or performance issues by narrowing the cause to either be on the server or off the server (as well as insight into how the server reacts under high usage).
All data is collected on a central server which also serves the web interface. Currently this server is Mon141. All servers have firewall rules opening the relevant ports that are needed so the central server can communicate with all the clients and clients can send metric information to the central server.
Adding New ServersEdit
This is done automatically on the first puppet run as Grafana configuration is handled in the base module. The central server can be replaced by adding
role::grafana to the server in site.pp.