akkoma/lib/pleroma/web/telemetry.ex

137 lines
4.3 KiB
Elixir
Raw Normal View History

2022-11-12 10:14:16 +00:00
defmodule Pleroma.Web.Telemetry do
use Supervisor
import Telemetry.Metrics
2022-11-12 16:13:39 +00:00
alias Pleroma.Stats
2022-11-12 10:14:16 +00:00
def start_link(arg) do
Supervisor.start_link(__MODULE__, arg, name: __MODULE__)
end
@impl true
def init(_arg) do
children = [
{:telemetry_poller, measurements: periodic_measurements(), period: 10_000},
2022-12-15 02:02:07 +00:00
{TelemetryMetricsPrometheus, metrics: prometheus_metrics(), plug_cowboy_opts: [ip: {127, 0, 0, 1}]}
2022-11-12 10:14:16 +00:00
]
Supervisor.init(children, strategy: :one_for_one)
end
2022-12-15 02:02:07 +00:00
@doc """
A seperate set of metrics for distributions because phoenix dashboard does NOT handle
them well
"""
defp distribution_metrics do
2022-11-12 10:14:16 +00:00
[
distribution(
"phoenix.router_dispatch.stop.duration",
# event_name: [:pleroma, :repo, :query, :total_time],
measurement: :duration,
unit: {:native, :second},
tags: [:route],
reporter_options: [
2022-11-12 10:23:44 +00:00
buckets: [0.1, 0.2, 0.5, 1, 2.5, 5, 10, 25, 50, 100, 250, 500, 1000]
2022-11-12 10:14:16 +00:00
]
),
# Database Time Metrics
distribution(
"pleroma.repo.query.total_time",
# event_name: [:pleroma, :repo, :query, :total_time],
measurement: :total_time,
unit: {:native, :millisecond},
reporter_options: [
2022-11-12 10:23:44 +00:00
buckets: [0.1, 0.2, 0.5, 1, 2.5, 5, 10, 25, 50, 100, 250, 500, 1000]
2022-11-12 10:14:16 +00:00
]
),
distribution(
"pleroma.repo.query.queue_time",
# event_name: [:pleroma, :repo, :query, :total_time],
measurement: :queue_time,
unit: {:native, :millisecond},
reporter_options: [
buckets: [0.01, 0.025, 0.05, 0.1, 0.2, 0.5, 1, 2.5, 5, 10]
]
),
distribution(
2022-12-15 02:02:07 +00:00
"oban_job_exception",
event_name: [:oban, :job, :exception],
2022-11-12 10:14:16 +00:00
measurement: :duration,
2022-11-12 10:54:35 +00:00
tags: [:worker],
tag_values: fn tags -> Map.put(tags, :worker, tags.job.worker) end,
2022-11-12 10:14:16 +00:00
unit: {:native, :second},
reporter_options: [
buckets: [0.01, 0.025, 0.05, 0.1, 0.2, 0.5, 1, 2.5, 5, 10]
]
2022-11-12 11:42:53 +00:00
),
distribution(
2022-12-15 02:02:07 +00:00
"tesla_request_completed",
event_name: [:tesla, :request, :stop],
2022-11-12 11:42:53 +00:00
measurement: :duration,
2022-12-15 02:02:07 +00:00
tags: [:response_code],
tag_values: fn tags -> Map.put(tags, :response_code, tags.env.status) end,
2022-11-12 11:42:53 +00:00
unit: {:native, :second},
reporter_options: [
buckets: [0.01, 0.025, 0.05, 0.1, 0.2, 0.5, 1, 2.5, 5, 10]
]
2022-11-12 15:11:38 +00:00
),
distribution(
2022-12-15 02:02:07 +00:00
"oban_job_completion",
event_name: [:oban, :job, :stop],
2022-11-12 15:11:38 +00:00
measurement: :duration,
2022-12-15 02:02:07 +00:00
tags: [:worker],
tag_values: fn tags -> Map.put(tags, :worker, tags.job.worker) end,
2022-11-12 15:11:38 +00:00
unit: {:native, :second},
reporter_options: [
buckets: [0.01, 0.025, 0.05, 0.1, 0.2, 0.5, 1, 2.5, 5, 10]
]
2022-12-15 02:02:07 +00:00
)
]
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}
2022-11-12 16:13:39 +00:00
),
2022-12-15 02:02:07 +00:00
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"),
2022-11-12 16:13:39 +00:00
last_value("pleroma.local_users.total"),
last_value("pleroma.domains.total"),
last_value("pleroma.local_statuses.total")
2022-11-12 10:14:16 +00:00
]
end
2022-12-15 02:02:07 +00:00
def prometheus_metrics, do: summary_metrics() ++ distribution_metrics()
def live_dashboard_metrics, do: summary_metrics()
2022-11-12 10:14:16 +00:00
defp periodic_measurements do
2022-11-12 16:13:39 +00:00
[
{__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}, %{})
2022-11-12 10:14:16 +00:00
end
end