File

net/server.lua @ 11523:5f15ab7c6ae5

Statistics: Rewrite statistics backends to use OpenMetrics The metric subsystem of Prosody has had some shortcomings from the perspective of the current state-of-the-art in metric observability. The OpenMetrics standard [0] is a formalization of the data model (and serialization format) of the well-known and widely-used Prometheus [1] software stack. The previous stats subsystem of Prosody did not map well to that format (see e.g. [2] and [3]); the key reason is that it was trying to do too much math on its own ([2]) while lacking first-class support for "families" of metrics ([3]) and structured metric metadata (despite the `extra` argument to metrics, there was no standard way of representing common things like "tags" or "labels"). Even though OpenMetrics has grown from the Prometheus world of monitoring, it maps well to other popular monitoring stacks such as: - InfluxDB (labels can be mapped to tags and fields as necessary) - Carbon/Graphite (labels can be attached to the metric name with dot-separation) - StatsD (see graphite when assuming that graphite is used as backend, which is the default) The util.statsd module has been ported to use the OpenMetrics model as a proof of concept. An implementation which exposes the util.statistics backend data as Prometheus metrics is ready for publishing in prosody-modules (most likely as mod_openmetrics_prometheus to avoid breaking existing 0.11 deployments). At the same time, the previous measure()-based API had one major advantage: It is really simple and easy to use without requiring lots of knowledge about OpenMetrics or similar concepts. For that reason as well as compatibility with existing code, it is preserved and may even be extended in the future. However, code relying on the `stats-updated` event as well as `get_stats` from `statsmanager` will break because the data model has changed completely; in case of `stats-updated`, the code will simply not run (as the event was renamed in order to avoid conflicts); the `get_stats` function has been removed completely (so it will cause a traceback when it is attempted to be used). Note that the measure_*_event methods have been removed from the module API. I was unable to find any uses or documentation and thus deemed they should not be ported. Re-implementation is possible when necessary. [0]: https://openmetrics.io/ [1]: https://prometheus.io/ [2]: #959 [3]: #960
author Jonas Schäfer <jonas@wielicki.name>
date Sun, 18 Apr 2021 11:47:41 +0200
parent 10853:0107ed6744aa
child 11762:54530085dffe
line wrap: on
line source

-- Prosody IM
-- Copyright (C) 2008-2010 Matthew Wild
-- Copyright (C) 2008-2010 Waqas Hussain
--
-- This project is MIT/X11 licensed. Please see the
-- COPYING file in the source package for more information.
--

if not (prosody and prosody.config_loaded) then
	-- This module only supports loading inside Prosody, outside Prosody
	-- you should directly require net.server_select or server_event, etc.
	error(debug.traceback("Loading outside Prosody or Prosody not yet initialized"), 0);
end

local log = require "util.logger".init("net.server");

local have_util_poll = pcall(require, "util.poll");
local default_backend = have_util_poll and "epoll" or "select";

local server_type = require "core.configmanager".get("*", "network_backend") or default_backend;

if require "core.configmanager".get("*", "use_libevent") then
	server_type = "event";
end

if server_type == "event" then
	if not pcall(require, "luaevent.core") then
		log("error", "libevent not found, falling back to select()");
		server_type = "select"
	end
end

local server;
local set_config;
if server_type == "event" then
	server = require "net.server_event";

	local defaults = {};
	for k,v in pairs(server.cfg) do
		defaults[k] = v;
	end
	function set_config(settings)
		local event_settings = {
			ACCEPT_DELAY = settings.accept_retry_interval;
			ACCEPT_QUEUE = settings.tcp_backlog;
			CLEAR_DELAY = settings.event_clear_interval;
			CONNECT_TIMEOUT = settings.connect_timeout;
			DEBUG = settings.debug;
			HANDSHAKE_TIMEOUT = settings.ssl_handshake_timeout;
			MAX_CONNECTIONS = settings.max_connections;
			MAX_HANDSHAKE_ATTEMPTS = settings.max_ssl_handshake_roundtrips;
			MAX_READ_LENGTH = settings.max_receive_buffer_size;
			MAX_SEND_LENGTH = settings.max_send_buffer_size;
			READ_TIMEOUT = settings.read_timeout;
			WRITE_TIMEOUT = settings.send_timeout;
		};

		for k,default in pairs(defaults) do
			server.cfg[k] = event_settings[k] or default;
		end
	end
elseif server_type == "select" then
	server = require "net.server_select";

	local defaults = {};
	for k,v in pairs(server.getsettings()) do
		defaults[k] = v;
	end
	function set_config(settings)
		local select_settings = {};
		for k,default in pairs(defaults) do
			select_settings[k] = settings[k] or default;
		end
		server.changesettings(select_settings);
	end
else
	server = require("net.server_"..server_type);
	set_config = server.set_config;
	if not server.get_backend then
		function server.get_backend()
			return server_type;
		end
	end
end

-- If server.hook_signal exists, replace signal.signal()
local has_signal, signal = pcall(require, "util.signal");
if has_signal then
	if server.hook_signal then
		function signal.signal(signal_id, handler)
			if type(signal_id) == "string" then
				signal_id = signal[signal_id:upper()];
			end
			if type(signal_id) ~= "number" then
				return false, "invalid-signal";
			end
			return server.hook_signal(signal_id, handler);
		end
	else
		server.hook_signal = signal.signal;
	end
else
	if not server.hook_signal then
		server.hook_signal = function()
			return false, "signal hooking not supported"
		end
	end
end

if prosody and set_config then
	local config_get = require "core.configmanager".get;
	local function load_config()
		local settings = config_get("*", "network_settings") or {};
		return set_config(settings);
	end
	load_config();
	prosody.events.add_handler("config-reloaded", load_config);
end

-- require "net.server" shall now forever return this,
-- ie. server_select or server_event as chosen above.
return server;