ddtrace 0.40.0 → 0.50.0

data/lib/ddtrace/configuration.rb

@@ -1,5 +1,4 @@
 require 'forwardable'
-
 require 'ddtrace/configuration/pin_setup'
 require 'ddtrace/configuration/settings'
 require 'ddtrace/configuration/components'
@@ -9,6 +8,33 @@ module Datadog
   module Configuration
     extend Forwardable
 
+    # Used to ensure that @components initialization/reconfiguration is performed one-at-a-time, by a single thread.
+    #
+    # This is important because components can end up being accessed from multiple application threads (for instance on
+    # a threaded webserver), and we don't want their initialization to clash (for instance, starting two profilers...).
+    #
+    # Note that a Mutex **IS NOT** reentrant: the same thread cannot grab the same Mutex more than once.
+    # This means below we are careful not to nest calls to methods that would trigger initialization and grab the lock.
+    #
+    # Every method that directly or indirectly mutates @components should be holding the lock (through
+    # #safely_synchronize) while doing so.
+    COMPONENTS_WRITE_LOCK = Mutex.new
+    private_constant :COMPONENTS_WRITE_LOCK
+
+    # We use a separate lock when reading the @components, so that they continue to be accessible during reconfiguration.
+    # This was needed because we ran into several issues where we still needed to read the old
+    # components while the COMPONENTS_WRITE_LOCK was being held (see https://github.com/DataDog/dd-trace-rb/pull/1387
+    # and https://github.com/DataDog/dd-trace-rb/pull/1373#issuecomment-799593022 ).
+    #
+    # Technically on MRI we could get away without this lock, but on non-MRI Rubies, we may run into issues because
+    # we fall into the "UnsafeDCLFactory" case of https://shipilev.net/blog/2014/safe-public-construction/ .
+    # Specifically, on JRuby reads from the @components do NOT have volatile semantics, and on TruffleRuby they do
+    # BUT just as an implementation detail, see https://github.com/jruby/jruby/wiki/Concurrency-in-jruby#volatility and
+    # https://github.com/DataDog/dd-trace-rb/pull/1329#issuecomment-776750377 .
+    # Concurrency is hard.
+    COMPONENTS_READ_LOCK = Mutex.new
+    private_constant :COMPONENTS_READ_LOCK
+
     attr_writer :configuration
 
     def configuration
@@ -19,13 +45,15 @@ module Datadog
       if target.is_a?(Settings)
         yield(target) if block_given?
 
-        # Build immutable components from settings
-        @components ||= nil
-        @components = if @components
-                        replace_components!(target, @components)
-                      else
-                        build_components(target)
-                      end
+        safely_synchronize do |write_components|
+          write_components.call(
+            if components?
+              replace_components!(target, @components)
+            else
+              build_components(target)
+            end
+          )
+        end
 
         target
       else
@@ -36,37 +64,91 @@ module Datadog
     def_delegators \
       :components,
       :health_metrics,
+      :profiler,
       :runtime_metrics,
       :tracer
 
     def logger
-      if instance_variable_defined?(:@components) && @components
+      # avoid initializing components if they didn't already exist
+      current_components = components(allow_initialization: false)
+
+      if current_components
         @temp_logger = nil
-        components.logger
+        current_components.logger
       else
-        # Use default logger without initializing components.
-        # This prevents recursive loops while initializing.
-        # e.g. Get logger --> Build components --> Log message --> Repeat...
-        @temp_logger ||= begin
-          logger = configuration.logger.instance || Datadog::Logger.new(STDOUT)
-          logger.level = configuration.diagnostics.debug ? ::Logger::DEBUG : configuration.logger.level
-          logger
-        end
+        logger_without_components
       end
     end
 
+    # Gracefully shuts down all components.
+    #
+    # Components will still respond to method calls as usual,
+    # but might not internally perform their work after shutdown.
+    #
+    # This avoids errors being raised across the host application
+    # during shutdown, while allowing for graceful decommission of resources.
+    #
+    # Components won't be automatically reinitialized after a shutdown.
     def shutdown!
-      components.shutdown! if instance_variable_defined?(:@components) && @components
+      safely_synchronize do
+        @components.shutdown! if components?
+      end
     end
 
     protected
 
-    def components
-      @components ||= build_components(configuration)
+    def components(allow_initialization: true)
+      current_components = COMPONENTS_READ_LOCK.synchronize { defined?(@components) && @components }
+      return current_components if current_components || !allow_initialization
+
+      safely_synchronize do |write_components|
+        (defined?(@components) && @components) || write_components.call(build_components(configuration))
+      end
     end
 
     private
 
+    # Gracefully shuts down the tracer and disposes of component references,
+    # allowing execution to start anew.
+    #
+    # In contrast with +#shutdown!+, components will be automatically
+    # reinitialized after a reset.
+    #
+    # Used internally to ensure a clean environment between test runs.
+    def reset!
+      safely_synchronize do |write_components|
+        @components.shutdown! if components?
+        write_components.call(nil)
+        configuration.reset!
+      end
+    end
+
+    def safely_synchronize
+      # Writes to @components should only happen through this proc. Because this proc is only accessible to callers of
+      # safely_synchronize, this forces all writers to go through this method.
+      write_components = proc do |new_value|
+        COMPONENTS_READ_LOCK.synchronize { @components = new_value }
+      end
+
+      COMPONENTS_WRITE_LOCK.synchronize do
+        begin
+          yield write_components
+        rescue ThreadError => e
+          logger_without_components.error(
+            'Detected deadlock during ddtrace initialization. ' \
+            'Please report this at https://github.com/DataDog/dd-trace-rb/blob/master/CONTRIBUTING.md#found-a-bug' \
+            "\n\tSource:\n\t#{e.backtrace.join("\n\t")}"
+          )
+          nil
+        end
+      end
+    end
+
+    def components?
+      # This does not need to grab the COMPONENTS_READ_LOCK because it's not returning the components
+      (defined?(@components) && @components) != nil
+    end
+
     def build_components(settings)
       components = Components.new(settings)
       components.startup!(settings)
@@ -80,5 +162,15 @@ module Datadog
       components.startup!(settings)
       components
     end
+
+    def logger_without_components
+      # Use default logger without initializing components.
+      # This enables logging during initialization, otherwise we'd run into deadlocks.
+      @temp_logger ||= begin
+        logger = configuration.logger.instance || Datadog::Logger.new($stdout)
+        logger.level = configuration.diagnostics.debug ? ::Logger::DEBUG : configuration.logger.level
+        logger
+      end
+    end
   end
 end

data/lib/ddtrace/configuration/agent_settings_resolver.rb

@@ -0,0 +1,230 @@
+require 'uri'
+
+require 'ddtrace/ext/transport'
+require 'ddtrace/configuration/settings'
+
+module Datadog
+  module Configuration
+    # This class unifies all the different ways that users can configure how we talk to the agent.
+    #
+    # It has quite a lot of complexity, but this complexity just reflects the actual complexity we have around our
+    # configuration today. E.g., this is just all of the complexity regarding agent settings gathered together in a
+    # single place. As we deprecate more and more of the different ways that these things can be configured,
+    # this class will reflect that simplification as well.
+    #
+    # Whenever there is a conflict (different configurations are provided in different orders), it MUST warn the users
+    # about it and pick a value based on the following priority: code > environment variable > defaults.
+    #
+    # rubocop:disable Metrics/ClassLength
+    class AgentSettingsResolver
+      AgentSettings = \
+        Struct.new(
+          :ssl,
+          :hostname,
+          :port,
+          :timeout_seconds,
+          :deprecated_for_removal_transport_configuration_proc,
+          :deprecated_for_removal_transport_configuration_options
+        ) do
+          def initialize(
+            ssl:,
+            hostname:,
+            port:,
+            timeout_seconds:,
+            deprecated_for_removal_transport_configuration_proc:,
+            deprecated_for_removal_transport_configuration_options:
+          )
+            super(ssl, hostname, port, timeout_seconds, deprecated_for_removal_transport_configuration_proc, \
+              deprecated_for_removal_transport_configuration_options)
+            freeze
+          end
+        end
+
+      def self.call(settings, logger: Datadog.logger)
+        new(settings, logger: logger).send(:call)
+      end
+
+      private
+
+      attr_reader \
+        :logger,
+        :settings
+
+      def initialize(settings, logger: Datadog.logger)
+        @settings = settings
+        @logger = logger
+      end
+
+      def call
+        AgentSettings.new(
+          ssl: ssl?,
+          hostname: hostname,
+          port: port,
+          timeout_seconds: timeout_seconds,
+          # NOTE: When provided, the deprecated_for_removal_transport_configuration_proc can override all
+          # values above (ssl, hostname, port, timeout), or even make them irrelevant (by using an unix socket or
+          # enabling test mode instead).
+          # That is the main reason why it is deprecated -- it's an opaque function that may set a bunch of settings
+          # that we know nothing of until we actually call it.
+          deprecated_for_removal_transport_configuration_proc: deprecated_for_removal_transport_configuration_proc,
+          deprecated_for_removal_transport_configuration_options: deprecated_for_removal_transport_configuration_options
+        )
+      end
+
+      def hostname
+        pick_from(
+          configurations_in_priority_order: [
+            DetectedConfiguration.new(
+              friendly_name: "'c.tracer.hostname'",
+              value: settings.tracer.hostname
+            ),
+            DetectedConfiguration.new(
+              friendly_name: "#{Datadog::Ext::Transport::HTTP::ENV_DEFAULT_URL} environment variable",
+              value: parsed_url && parsed_url.hostname
+            ),
+            DetectedConfiguration.new(
+              friendly_name: "#{Datadog::Ext::Transport::HTTP::ENV_DEFAULT_HOST} environment variable",
+              value: ENV[Datadog::Ext::Transport::HTTP::ENV_DEFAULT_HOST]
+            )
+          ],
+          or_use_default: Datadog::Ext::Transport::HTTP::DEFAULT_HOST
+        )
+      end
+
+      def port
+        port_from_env = ENV[Datadog::Ext::Transport::HTTP::ENV_DEFAULT_PORT]
+        parsed_port_from_env =
+          if port_from_env
+            begin
+              Integer(port_from_env)
+            rescue ArgumentError
+              log_warning(
+                "Invalid value for #{Datadog::Ext::Transport::HTTP::ENV_DEFAULT_PORT} environment variable " \
+                "('#{port_from_env}'). Ignoring this configuration."
+              )
+            end
+          end
+
+        pick_from(
+          configurations_in_priority_order: [
+            DetectedConfiguration.new(
+              friendly_name: '"c.tracer.port"',
+              value: settings.tracer.port
+            ),
+            DetectedConfiguration.new(
+              friendly_name: "#{Datadog::Ext::Transport::HTTP::ENV_DEFAULT_URL} environment variable",
+              value: parsed_url && parsed_url.port
+            ),
+            DetectedConfiguration.new(
+              friendly_name: "#{Datadog::Ext::Transport::HTTP::ENV_DEFAULT_PORT} environment variable",
+              value: parsed_port_from_env
+            )
+          ],
+          or_use_default: Datadog::Ext::Transport::HTTP::DEFAULT_PORT
+        )
+      end
+
+      def ssl?
+        !parsed_url.nil? && parsed_url.scheme == 'https'
+      end
+
+      def timeout_seconds
+        Datadog::Ext::Transport::HTTP::DEFAULT_TIMEOUT_SECONDS
+      end
+
+      def deprecated_for_removal_transport_configuration_proc
+        settings.tracer.transport_options if settings.tracer.transport_options.is_a?(Proc)
+      end
+
+      def deprecated_for_removal_transport_configuration_options
+        options = settings.tracer.transport_options
+
+        if options.is_a?(Hash) && !options.empty?
+          log_warning(
+            'Configuring the tracer via a c.tracer.transport_options hash is deprecated for removal in a future ' \
+            "ddtrace version (c.tracer.transport_options contained '#{options.inspect}')."
+          )
+
+          options
+        end
+      end
+
+      def parsed_url
+        return @parsed_url if defined?(@parsed_url)
+
+        @parsed_url =
+          if unparsed_url_from_env
+            parsed = URI.parse(unparsed_url_from_env)
+
+            if %w[http https].include?(parsed.scheme)
+              parsed
+            else
+              log_warning(
+                "Invalid URI scheme '#{parsed.scheme}' for #{Datadog::Ext::Transport::HTTP::ENV_DEFAULT_URL} " \
+                "environment variable ('#{unparsed_url_from_env}'). " \
+                "Ignoring the contents of #{Datadog::Ext::Transport::HTTP::ENV_DEFAULT_URL}."
+              )
+
+              nil
+            end
+          end
+      end
+
+      # NOTE: This should only be used AFTER parsing, via `#parsed_url`. The only other use-case where this can be used
+      # directly without parsing, is when displaying in warning messages, to show users what it actually contains.
+      def unparsed_url_from_env
+        @unparsed_url_from_env ||= ENV[Datadog::Ext::Transport::HTTP::ENV_DEFAULT_URL]
+      end
+
+      def pick_from(configurations_in_priority_order:, or_use_default:)
+        detected_configurations_in_priority_order = configurations_in_priority_order.select(&:value?)
+
+        if detected_configurations_in_priority_order.any?
+          warn_if_configuration_mismatch(detected_configurations_in_priority_order)
+
+          # The configurations are listed in priority, so we only need to look at the first; if there's more than
+          # one, we emit a warning above
+          detected_configurations_in_priority_order.first.value
+        else
+          or_use_default
+        end
+      end
+
+      def warn_if_configuration_mismatch(detected_configurations_in_priority_order)
+        return unless detected_configurations_in_priority_order.map(&:value).uniq.size > 1
+
+        log_warning(
+          'Configuration mismatch: values differ between ' \
+          "#{detected_configurations_in_priority_order
+            .map { |config| "#{config.friendly_name} ('#{config.value}')" }.join(' and ')}" \
+          ". Using '#{detected_configurations_in_priority_order.first.value}'."
+        )
+      end
+
+      def log_warning(message)
+        logger.warn(message) if logger
+      end
+
+      DetectedConfiguration = Struct.new(:friendly_name, :value) do
+        def initialize(friendly_name:, value:)
+          super(friendly_name, value)
+          freeze
+        end
+
+        def value?
+          !value.nil?
+        end
+      end
+      private_constant :DetectedConfiguration
+
+      # NOTE: Due to... legacy reasons... Some classes like having an `AgentSettings` instance to fall back to.
+      # Because we generate this instance with an empty instance of `Settings`, the resulting `AgentSettings` below
+      # represents only settings specified via environment variables + the usual defaults.
+      #
+      # YOU DO NOT WANT TO USE THE BELOW INSTANCE ON ANY NEWLY WRITTEN CODE, as it ignores any settings specified
+      # by users via `Datadog.configure`.
+      ENVIRONMENT_AGENT_SETTINGS = call(Settings.new, logger: nil)
+    end
+    # rubocop:enable Metrics/ClassLength
+  end
+end

data/lib/ddtrace/configuration/base.rb

@@ -6,12 +6,12 @@ module Datadog
     # Basic configuration behavior
     module Base
       def self.included(base)
-        base.send(:extend, Datadog::Environment::Helpers)
-        base.send(:include, Datadog::Environment::Helpers)
-        base.send(:include, Options)
+        base.extend(Datadog::Environment::Helpers)
+        base.include(Datadog::Environment::Helpers)
+        base.include(Options)
 
-        base.send(:extend, ClassMethods)
-        base.send(:include, InstanceMethods)
+        base.extend(ClassMethods)
+        base.include(InstanceMethods)
       end
 
       # Class methods for configuration
@@ -37,7 +37,7 @@ module Datadog
 
         def new_settings_class(&block)
           Class.new { include Datadog::Configuration::Base }.tap do |klass|
-            klass.instance_eval(&block) if block_given?
+            klass.instance_eval(&block) if block
           end
         end
       end
@@ -53,10 +53,7 @@ module Datadog
           ordering = self.class.options.dependency_order
           sorted_opts = opts.sort_by do |name, _value|
             ordering.index(name) || (ordering.length + 1)
-          end
-
-          # Ruby 2.0 doesn't support Array#to_h
-          sorted_opts = Hash[*sorted_opts.flatten]
+          end.to_h
 
           # Apply options in sort order
           sorted_opts.each do |name, value|