Merge pull request 'metrics' (#375) from stats into develop

[akkoma] / lib / pleroma / web / telemetry.ex

diff --git a/lib/pleroma/web/telemetry.ex b/lib/pleroma/web/telemetry.ex
index d9c54c28f595ec3bf8c60a8b4ce782203e148568..5b01ee14deeb97218b0ea168c78f90153c1a736e 100644 (file)
--- a/lib/pleroma/web/telemetry.ex
+++ b/lib/pleroma/web/telemetry.ex
@@ -1,6 +1,7 @@
 defmodule Pleroma.Web.Telemetry do
   use Supervisor
   import Telemetry.Metrics
+  alias Pleroma.Stats
 
   def start_link(arg) do
     Supervisor.start_link(__MODULE__, arg, name: __MODULE__)
@@ -9,27 +10,16 @@ defmodule Pleroma.Web.Telemetry do
   @impl true
   def init(_arg) do
     children = [
-      # Telemetry poller will execute the given period measurements
-      # every 10_000ms. Learn more here: https://hexdocs.pm/telemetry_metrics
       {:telemetry_poller, measurements: periodic_measurements(), period: 10_000},
-      # Add reporters as children of your supervision tree.
-      # {Telemetry.Metrics.ConsoleReporter, metrics: metrics()},
-      {TelemetryMetricsPrometheus, metrics: metrics(), plug_cowboy_opts: [ip: {127, 0, 0, 1}]}
+      {TelemetryMetricsPrometheus.Core, metrics: prometheus_metrics()}
     ]
 
     Supervisor.init(children, strategy: :one_for_one)
   end
 
-  def metrics do
+  # A seperate set of metrics for distributions because phoenix dashboard does NOT handle them well
+  defp distribution_metrics do
     [
-      # Phoenix Metrics
-      summary("phoenix.endpoint.stop.duration",
-        unit: {:native, :millisecond}
-      ),
-      summary("phoenix.router_dispatch.stop.duration",
-        tags: [:route],
-        unit: {:native, :millisecond}
-      ),
       distribution(
         "phoenix.router_dispatch.stop.duration",
         # event_name: [:pleroma, :repo, :query, :total_time],
@@ -60,20 +50,9 @@ defmodule Pleroma.Web.Telemetry do
           buckets: [0.01, 0.025, 0.05, 0.1, 0.2, 0.5, 1, 2.5, 5, 10]
         ]
       ),
-      summary("pleroma.repo.query.total_time", unit: {:native, :millisecond}),
-      summary("pleroma.repo.query.decode_time", unit: {:native, :millisecond}),
-      summary("pleroma.repo.query.query_time", unit: {:native, :millisecond}),
-      summary("pleroma.repo.query.queue_time", unit: {:native, :millisecond}),
-      summary("pleroma.repo.query.idle_time", 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"),
       distribution(
-        "oban_job_completion",
-        event_name: [:oban, :job, :stop],
+        "oban_job_exception",
+        event_name: [:oban, :job, :exception],
         measurement: :duration,
         tags: [:worker],
         tag_values: fn tags -> Map.put(tags, :worker, tags.job.worker) end,
@@ -83,8 +62,19 @@ defmodule Pleroma.Web.Telemetry do
         ]
       ),
       distribution(
-        "oban_job_exception",
-        event_name: [:oban, :job, :exception],
+        "tesla_request_completed",
+        event_name: [:tesla, :request, :stop],
+        measurement: :duration,
+        tags: [:response_code],
+        tag_values: fn tags -> Map.put(tags, :response_code, tags.env.status) end,
+        unit: {:native, :second},
+        reporter_options: [
+          buckets: [0.01, 0.025, 0.05, 0.1, 0.2, 0.5, 1, 2.5, 5, 10]
+        ]
+      ),
+      distribution(
+        "oban_job_completion",
+        event_name: [:oban, :job, :stop],
         measurement: :duration,
         tags: [:worker],
         tag_values: fn tags -> Map.put(tags, :worker, tags.job.worker) end,
@@ -96,7 +86,46 @@ defmodule Pleroma.Web.Telemetry do
     ]
   end
 
+  defp summary_metrics do
+    [
+      # Phoenix Metrics
+      summary("phoenix.endpoint.stop.duration",
+        unit: {:native, :millisecond}
+      ),
+      summary("phoenix.router_dispatch.stop.duration",
+        tags: [:route],
+        unit: {:native, :millisecond}
+      ),
+      summary("pleroma.repo.query.total_time", unit: {:native, :millisecond}),
+      summary("pleroma.repo.query.decode_time", unit: {:native, :millisecond}),
+      summary("pleroma.repo.query.query_time", unit: {:native, :millisecond}),
+      summary("pleroma.repo.query.queue_time", unit: {:native, :millisecond}),
+      summary("pleroma.repo.query.idle_time", 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"),
+      last_value("pleroma.local_users.total"),
+      last_value("pleroma.domains.total"),
+      last_value("pleroma.local_statuses.total")
+    ]
+  end
+
+  def prometheus_metrics, do: summary_metrics() ++ distribution_metrics()
+  def live_dashboard_metrics, do: summary_metrics()
+
   defp periodic_measurements do
-    []
+    [
+      {__MODULE__, :instance_stats, []}
+    ]
+  end
+
+  def instance_stats do
+    stats = Stats.get_stats()
+    :telemetry.execute([:pleroma, :local_users], %{total: stats.user_count}, %{})
+    :telemetry.execute([:pleroma, :domains], %{total: stats.domain_count}, %{})
+    :telemetry.execute([:pleroma, :local_statuses], %{total: stats.status_count}, %{})
   end
 end