defmodule Observer.Web.Telemetry do use Supervisor import Telemetry.Metrics import ObserverWeb.Macros def start_link(arg) do Supervisor.start_link(__MODULE__, arg, name: __MODULE__) end @impl true def init(_arg) do # Telemetry poller will execute the given period measurements # every 5_000ms. Learn more here: https://hexdocs.pm/telemetry_metrics children = add_telemetry_poller() ++ [ # Add reporters as children of your supervision tree. {ObserverWeb.Telemetry.Consumer, metrics: metrics()} ] Supervisor.init(children, strategy: :one_for_one) end if_not_test do defp add_telemetry_poller, do: [ {:telemetry_poller, measurements: periodic_measurements(), period: 5_000} ] else defp add_telemetry_poller, do: [] end def metrics do [ # Phoenix Metrics summary("phoenix.endpoint.start.system_time", unit: {:native, :millisecond} ), summary("phoenix.endpoint.stop.duration", unit: {:native, :millisecond} ), summary("phoenix.router_dispatch.start.system_time", tags: [:route], unit: {:native, :millisecond} ), summary("phoenix.router_dispatch.exception.duration", tags: [:route], unit: {:native, :millisecond} ), summary("phoenix.router_dispatch.stop.duration", tags: [:route], unit: {:native, :millisecond} ), summary("phoenix.socket_connected.duration", unit: {:native, :millisecond} ), summary("phoenix.channel_joined.duration", unit: {:native, :millisecond} ), summary("phoenix.channel_handled_in.duration", tags: [:event], unit: {:native, :millisecond} ), # VM Metrics summary("vm.memory.total", unit: {:byte, :kilobyte}), summary("vm.total_run_queue_lengths.total"), summary("vm.total_run_queue_lengths.cpu"), summary("vm.total_run_queue_lengths.io") ] end if_not_test do defp periodic_measurements do [ # A module, function and arguments to be invoked periodically. # This function must call :telemetry.execute/3 and a metric must be added above. {ObserverWeb.Telemetry.Producer.PhxLvSocket, :process_phoenix_liveview_sockets, []} ] end end end