ddtrace 0.53.0 → 1.0.0.beta1

data/lib/datadog/profiling/native_extension.rb

@@ -0,0 +1,40 @@
+# typed: true
+module Datadog
+  module Profiling
+    # This module contains classes and methods which are implemented using native code in the
+    # ext/ddtrace_profiling_native_extension folder, as well as some Ruby-level utilities that don't make sense to
+    # write using C
+    module NativeExtension
+      private_class_method def self.working?
+        native_working?
+      end
+
+      unless singleton_class.private_method_defined?(:native_working?)
+        private_class_method def self.native_working?
+          false
+        end
+      end
+
+      unless singleton_class.method_defined?(:clock_id_for)
+        def self.clock_id_for(_)
+          nil
+        end
+      end
+
+      def self.cpu_time_ns_for(thread)
+        clock_id =
+          begin
+            clock_id_for(thread)
+          rescue Errno::ESRCH
+            nil
+          end
+
+        begin
+          ::Process.clock_gettime(clock_id, :nanosecond) if clock_id
+        rescue Errno::EINVAL
+          nil
+        end
+      end
+    end
+  end
+end

data/lib/datadog/profiling/pprof/builder.rb

@@ -0,0 +1,126 @@
+# typed: true
+# frozen_string_literal: true
+
+require 'datadog/profiling/flush'
+require 'datadog/profiling/pprof/message_set'
+require 'datadog/profiling/pprof/string_table'
+require 'datadog/core/utils/time'
+
+module Datadog
+  module Profiling
+    module Pprof
+      # Accumulates profile data and produces a Perftools::Profiles::Profile
+      class Builder
+        DEFAULT_ENCODING = 'UTF-8'
+        DESC_FRAME_OMITTED = 'frame omitted'
+        DESC_FRAMES_OMITTED = 'frames omitted'
+
+        attr_reader \
+          :functions,
+          :locations,
+          :mappings,
+          :sample_types,
+          :samples,
+          :string_table
+
+        def initialize
+          @functions = MessageSet.new(1)
+          @locations = initialize_locations_hash
+          @mappings = MessageSet.new(1)
+          @sample_types = MessageSet.new
+          @samples = []
+          @string_table = StringTable.new
+
+          # Cache this proc, since it's pretty expensive to keep recreating it
+          @build_function = method(:build_function).to_proc
+        end
+
+        # The locations hash maps unique BacktraceLocation instances to their corresponding pprof Location objects;
+        # there's a 1:1 correspondence, since BacktraceLocations were already deduped
+        def initialize_locations_hash
+          sequence = Core::Utils::Sequence.new(1)
+          Hash.new do |locations_hash, backtrace_location|
+            locations_hash[backtrace_location] = build_location(sequence.next, backtrace_location)
+          end
+        end
+
+        def encode_profile(profile)
+          Perftools::Profiles::Profile.encode(profile).force_encoding(DEFAULT_ENCODING)
+        end
+
+        def build_profile(start:, finish:)
+          start_ns = Core::Utils::Time.as_utc_epoch_ns(start)
+          finish_ns = Core::Utils::Time.as_utc_epoch_ns(finish)
+
+          Perftools::Profiles::Profile.new(
+            sample_type: @sample_types.messages,
+            sample: @samples,
+            mapping: @mappings.messages,
+            location: @locations.values,
+            function: @functions.messages,
+            string_table: @string_table.strings,
+            time_nanos: start_ns,
+            duration_nanos: finish_ns - start_ns,
+          )
+        end
+
+        def build_value_type(type, unit)
+          Perftools::Profiles::ValueType.new(
+            type: @string_table.fetch(type),
+            unit: @string_table.fetch(unit)
+          )
+        end
+
+        def build_locations(backtrace_locations, length)
+          locations = backtrace_locations.collect { |backtrace_location| @locations[backtrace_location] }
+
+          omitted = length - backtrace_locations.length
+
+          # Add placeholder stack frame if frames were truncated
+          if omitted > 0
+            desc = omitted == 1 ? DESC_FRAME_OMITTED : DESC_FRAMES_OMITTED
+            locations << @locations[Profiling::BacktraceLocation.new('', 0, "#{omitted} #{desc}")]
+          end
+
+          locations
+        end
+
+        def build_location(id, backtrace_location)
+          Perftools::Profiles::Location.new(
+            id: id,
+            line: [build_line(
+              @functions.fetch(
+                backtrace_location.path,
+                backtrace_location.base_label,
+                &@build_function
+              ).id,
+              backtrace_location.lineno
+            )]
+          )
+        end
+
+        def build_line(function_id, line_number)
+          Perftools::Profiles::Line.new(
+            function_id: function_id,
+            line: line_number
+          )
+        end
+
+        def build_function(id, filename, function_name)
+          Perftools::Profiles::Function.new(
+            id: id,
+            name: @string_table.fetch(function_name),
+            filename: @string_table.fetch(filename)
+          )
+        end
+
+        def build_mapping(id, filename)
+          Perftools::Profiles::Mapping.new(
+            id: id,
+            filename: @string_table.fetch(filename)
+          )
+        end
+      end
+    end
+  end
+end

data/lib/datadog/profiling/pprof/converter.rb

@@ -0,0 +1,103 @@
+# typed: true
+require 'datadog/profiling/ext'
+
+module Datadog
+  module Profiling
+    module Pprof
+      # Base class for converters that convert profiling events to pprof
+      class Converter
+        attr_reader \
+          :builder
+
+        # Override in child class to define sample types
+        # this converter uses when building samples.
+        def self.sample_value_types
+          raise NotImplementedError
+        end
+
+        def initialize(builder, sample_type_mappings)
+          @builder = builder
+          @sample_type_mappings = sample_type_mappings
+        end
+
+        def group_events(events)
+          # Event grouping in format:
+          # [key, EventGroup]
+          event_groups = {}
+
+          # Aggregate each event into a group
+          # with identical properties, but different values.
+          events.each do |event|
+            key = yield(event)
+            values = build_event_values(event)
+
+            unless key.nil?
+              if event_groups.key?(key)
+                # Update existing group from event
+                update_group(event_groups[key], event, values)
+              else
+                # Add new group
+                event_groups[key] = new_group(event, values)
+              end
+            end
+          end
+
+          event_groups
+        end
+
+        def add_events!(events)
+          raise NotImplementedError
+        end
+
+        def sample_value_index(type)
+          index = @sample_type_mappings[type]
+          raise UnknownSampleTypeMappingError, type unless index
+
+          index
+        end
+
+        def build_event_values(event)
+          # Build a value array that matches the length of the sample types
+          # Populate all values with "no value" by default
+          Array.new(@sample_type_mappings.length, Profiling::Ext::Pprof::SAMPLE_VALUE_NO_VALUE)
+        end
+
+        def debug_statistics
+          # Empty; can be used by subclasses to report a string containing debug statistics to be logged
+        end
+
+        # Represents a grouped event
+        # 'sample' is an example event object from the group.
+        # 'values' is the summation of the group's sample values
+        EventGroup = Struct.new(:sample, :values)
+
+        # Error when the mapping of a sample type to value index is unknown
+        class UnknownSampleTypeMappingError < StandardError
+          attr_reader :type
+
+          def initialize(type)
+            @type = type
+          end
+
+          def message
+            "Mapping for sample value type '#{type}' to index is unknown."
+          end
+        end
+
+        protected
+
+        def new_group(event, values)
+          EventGroup.new(event, values)
+        end
+
+        def update_group(event_group, event, values)
+          # Update values for group
+          group_values = event_group.values
+          group_values.each_with_index do |group_value, i|
+            group_values[i] = group_value + values[i]
+          end
+        end
+      end
+    end
+  end
+end

data/lib/datadog/profiling/pprof/message_set.rb

@@ -0,0 +1,15 @@
+# typed: true
+require 'datadog/core/utils/object_set'
+
+module Datadog
+  module Profiling
+    module Pprof
+      # Acts as a unique dictionary of protobuf messages
+      class MessageSet < Core::Utils::ObjectSet
+        def messages
+          objects
+        end
+      end
+    end
+  end
+end

data/lib/datadog/profiling/pprof/pprof_pb.rb

@@ -0,0 +1,82 @@
+# typed: false
+# Generated by the protocol buffer compiler.  DO NOT EDIT!
+# source: lib/datadog/profiling/pprof/pprof.proto
+
+require 'google/protobuf'
+
+Google::Protobuf::DescriptorPool.generated_pool.build do
+  add_message "perftools.profiles.Profile" do
+    repeated :sample_type, :message, 1, "perftools.profiles.ValueType"
+    repeated :sample, :message, 2, "perftools.profiles.Sample"
+    repeated :mapping, :message, 3, "perftools.profiles.Mapping"
+    repeated :location, :message, 4, "perftools.profiles.Location"
+    repeated :function, :message, 5, "perftools.profiles.Function"
+    repeated :string_table, :string, 6
+    optional :drop_frames, :int64, 7
+    optional :keep_frames, :int64, 8
+    optional :time_nanos, :int64, 9
+    optional :duration_nanos, :int64, 10
+    optional :period_type, :message, 11, "perftools.profiles.ValueType"
+    optional :period, :int64, 12
+    repeated :comment, :int64, 13
+    optional :default_sample_type, :int64, 14
+  end
+  add_message "perftools.profiles.ValueType" do
+    optional :type, :int64, 1
+    optional :unit, :int64, 2
+  end
+  add_message "perftools.profiles.Sample" do
+    repeated :location_id, :uint64, 1
+    repeated :value, :int64, 2
+    repeated :label, :message, 3, "perftools.profiles.Label"
+  end
+  add_message "perftools.profiles.Label" do
+    optional :key, :int64, 1
+    optional :str, :int64, 2
+    optional :num, :int64, 3
+    optional :num_unit, :int64, 4
+  end
+  add_message "perftools.profiles.Mapping" do
+    optional :id, :uint64, 1
+    optional :memory_start, :uint64, 2
+    optional :memory_limit, :uint64, 3
+    optional :file_offset, :uint64, 4
+    optional :filename, :int64, 5
+    optional :build_id, :int64, 6
+    optional :has_functions, :bool, 7
+    optional :has_filenames, :bool, 8
+    optional :has_line_numbers, :bool, 9
+    optional :has_inline_frames, :bool, 10
+  end
+  add_message "perftools.profiles.Location" do
+    optional :id, :uint64, 1
+    optional :mapping_id, :uint64, 2
+    optional :address, :uint64, 3
+    repeated :line, :message, 4, "perftools.profiles.Line"
+    optional :is_folded, :bool, 5
+  end
+  add_message "perftools.profiles.Line" do
+    optional :function_id, :uint64, 1
+    optional :line, :int64, 2
+  end
+  add_message "perftools.profiles.Function" do
+    optional :id, :uint64, 1
+    optional :name, :int64, 2
+    optional :system_name, :int64, 3
+    optional :filename, :int64, 4
+    optional :start_line, :int64, 5
+  end
+end
+
+module Perftools
+  module Profiles
+    Profile = Google::Protobuf::DescriptorPool.generated_pool.lookup("perftools.profiles.Profile").msgclass
+    ValueType = Google::Protobuf::DescriptorPool.generated_pool.lookup("perftools.profiles.ValueType").msgclass
+    Sample = Google::Protobuf::DescriptorPool.generated_pool.lookup("perftools.profiles.Sample").msgclass
+    Label = Google::Protobuf::DescriptorPool.generated_pool.lookup("perftools.profiles.Label").msgclass
+    Mapping = Google::Protobuf::DescriptorPool.generated_pool.lookup("perftools.profiles.Mapping").msgclass
+    Location = Google::Protobuf::DescriptorPool.generated_pool.lookup("perftools.profiles.Location").msgclass
+    Line = Google::Protobuf::DescriptorPool.generated_pool.lookup("perftools.profiles.Line").msgclass
+    Function = Google::Protobuf::DescriptorPool.generated_pool.lookup("perftools.profiles.Function").msgclass
+  end
+end

data/lib/datadog/profiling/pprof/stack_sample.rb

@@ -0,0 +1,140 @@
+# typed: true
+require 'datadog/profiling/ext'
+require 'datadog/profiling/events/stack'
+require 'datadog/profiling/pprof/builder'
+require 'datadog/profiling/pprof/converter'
+
+module Datadog
+  module Profiling
+    module Pprof
+      # Builds a profile from a StackSample
+      #
+      # NOTE: This class may appear stateless but is in fact stateful; a new instance should be created for every
+      # encoded profile.
+      class StackSample < Converter
+        SAMPLE_TYPES = {
+          cpu_time_ns: [
+            Profiling::Ext::Pprof::VALUE_TYPE_CPU,
+            Profiling::Ext::Pprof::VALUE_UNIT_NANOSECONDS
+          ],
+          wall_time_ns: [
+            Profiling::Ext::Pprof::VALUE_TYPE_WALL,
+            Profiling::Ext::Pprof::VALUE_UNIT_NANOSECONDS
+          ]
+        }.freeze
+
+        def self.sample_value_types
+          SAMPLE_TYPES
+        end
+
+        def initialize(*_)
+          super
+
+          @most_recent_trace_samples = {}
+          @processed_unique_stacks = 0
+          @processed_with_trace = 0
+        end
+
+        def add_events!(stack_samples)
+          new_samples = build_samples(stack_samples)
+          builder.samples.concat(new_samples)
+        end
+
+        def stack_sample_group_key(stack_sample)
+          # We want to make sure we have the most recent sample for any trace.
+          # (This is done here to save an iteration over all samples.)
+          update_most_recent_trace_sample(stack_sample)
+
+          stack_sample.hash
+        end
+
+        # Track the most recent sample for each trace (identified by root span id)
+        def update_most_recent_trace_sample(stack_sample)
+          return unless stack_sample.root_span_id && stack_sample.trace_resource
+
+          # Update trace resource with most recent value
+          if (most_recent_trace_sample = @most_recent_trace_samples[stack_sample.root_span_id])
+            if most_recent_trace_sample.timestamp < stack_sample.timestamp
+              @most_recent_trace_samples[stack_sample.root_span_id] = stack_sample
+            end
+          else
+            # Add trace resource
+            @most_recent_trace_samples[stack_sample.root_span_id] = stack_sample
+          end
+        end
+
+        def build_samples(stack_samples)
+          groups = group_events(stack_samples, &method(:stack_sample_group_key))
+          groups.collect do |_group_key, group|
+            @processed_unique_stacks += 1
+            build_sample(group.sample, group.values)
+          end
+        end
+
+        def build_sample(stack_sample, values)
+          locations = builder.build_locations(
+            stack_sample.frames,
+            stack_sample.total_frame_count
+          )
+
+          Perftools::Profiles::Sample.new(
+            location_id: locations.collect { |location| location['id'.freeze] },
+            value: values,
+            label: build_sample_labels(stack_sample)
+          )
+        end
+
+        def build_event_values(stack_sample)
+          no_value = Profiling::Ext::Pprof::SAMPLE_VALUE_NO_VALUE
+          values = super(stack_sample)
+          values[sample_value_index(:cpu_time_ns)] = stack_sample.cpu_time_interval_ns || no_value
+          values[sample_value_index(:wall_time_ns)] = stack_sample.wall_time_interval_ns || no_value
+          values
+        end
+
+        def build_sample_labels(stack_sample)
+          labels = [
+            Perftools::Profiles::Label.new(
+              key: builder.string_table.fetch(Profiling::Ext::Pprof::LABEL_KEY_THREAD_ID),
+              str: builder.string_table.fetch(stack_sample.thread_id.to_s)
+            )
+          ]
+
+          root_span_id = stack_sample.root_span_id || 0
+          span_id = stack_sample.span_id || 0
+
+          if root_span_id != 0 && span_id != 0
+            @processed_with_trace += 1
+
+            labels << Perftools::Profiles::Label.new(
+              key: builder.string_table.fetch(Profiling::Ext::Pprof::LABEL_KEY_LOCAL_ROOT_SPAN_ID),
+              str: builder.string_table.fetch(root_span_id.to_s)
+            )
+
+            labels << Perftools::Profiles::Label.new(
+              key: builder.string_table.fetch(Profiling::Ext::Pprof::LABEL_KEY_SPAN_ID),
+              str: builder.string_table.fetch(span_id.to_s)
+            )
+
+            # Use most up-to-date trace resource, if available.
+            # Otherwise, use the trace resource provided.
+            trace_resource = @most_recent_trace_samples.fetch(stack_sample.root_span_id, stack_sample).trace_resource
+
+            if trace_resource && !trace_resource.empty?
+              labels << Perftools::Profiles::Label.new(
+                key: builder.string_table.fetch(Profiling::Ext::Pprof::LABEL_KEY_TRACE_ENDPOINT),
+                str: builder.string_table.fetch(trace_resource)
+              )
+            end
+          end
+
+          labels
+        end
+
+        def debug_statistics
+          "unique stacks: #{@processed_unique_stacks}, of which had active traces: #{@processed_with_trace}"
+        end
+      end
+    end
+  end
+end

data/lib/datadog/profiling/pprof/string_table.rb

@@ -0,0 +1,11 @@
+# typed: strict
+require 'datadog/core/utils/string_table'
+
+module Datadog
+  module Profiling
+    module Pprof
+      # Tracks strings and returns IDs
+      class StringTable < Core::Utils::StringTable; end
+    end
+  end
+end

data/lib/datadog/profiling/pprof/template.rb

@@ -0,0 +1,119 @@
+# typed: false
+require 'datadog/profiling/pprof/payload'
+require 'datadog/profiling/pprof/message_set'
+require 'datadog/profiling/pprof/builder'
+
+require 'datadog/profiling/events/stack'
+require 'datadog/profiling/pprof/stack_sample'
+
+module Datadog
+  module Profiling
+    module Pprof
+      # Converts a collection of profiling events into a Perftools::Profiles::Profile
+      class Template
+        DEFAULT_MAPPINGS = {
+          Events::StackSample => Pprof::StackSample
+        }.freeze
+
+        attr_reader \
+          :builder,
+          :converters,
+          :sample_type_mappings
+
+        def self.for_event_classes(event_classes)
+          # Build a map of event class --> converter class
+          mappings = event_classes.each_with_object({}) do |event_class, m|
+            converter_class = DEFAULT_MAPPINGS[event_class]
+            raise NoProfilingEventConversionError, event_class unless converter_class
+
+            m[event_class] = converter_class
+          end
+
+          new(mappings)
+        end
+
+        def initialize(mappings)
+          @builder = Builder.new
+          @converters = Hash.new { |_h, event_class| raise NoProfilingEventConversionError, event_class }
+          @sample_type_mappings = Hash.new { |_h, type| raise UnknownSampleTypeMappingError, type }
+
+          # Add default mapping
+          builder.mappings.fetch($PROGRAM_NAME, &builder.method(:build_mapping))
+
+          # Combine all sample types from each converter class
+          types = mappings.values.each_with_object({}) do |converter_class, t|
+            t.merge!(converter_class.sample_value_types)
+          end
+
+          # Build the sample types into sample type objects
+          types.each do |type_name, type_args|
+            index = nil
+
+            sample_type = builder.sample_types.fetch(*type_args) do |id, type, unit|
+              index = id
+              builder.build_value_type(type, unit)
+            end
+
+            # Create mapping between the type and index to which its assigned.
+            # Do this for faster lookup while building profile sample values.
+            sample_type_mappings[type_name] = index || builder.sample_types.messages.index(sample_type)
+          end
+
+          # Freeze them so they can't be modified.
+          # We don't want the number of sample types to vary between samples within the same profile.
+          builder.sample_types.freeze
+          sample_type_mappings.freeze
+
+          # Add converters
+          mappings.each do |event_class, converter_class|
+            converters[event_class] = converter_class.new(builder, sample_type_mappings)
+          end
+
+          converters.freeze
+        end
+
+        def add_events!(event_class, events)
+          converters[event_class].add_events!(events)
+        end
+
+        def debug_statistics
+          converters.values.map(&:debug_statistics).join(', ')
+        end
+
+        def to_pprof(start:, finish:)
+          profile = builder.build_profile(start: start, finish: finish)
+          data = builder.encode_profile(profile)
+          types = sample_type_mappings.keys
+
+          Payload.new(data, types)
+        end
+
+        # Error when an unknown event type is given to be converted
+        class NoProfilingEventConversionError < ArgumentError
+          attr_reader :type
+
+          def initialize(type)
+            @type = type
+          end
+
+          def message
+            "Profiling event type '#{type}' cannot be converted to pprof."
+          end
+        end
+
+        # Error when the mapping of a sample type to value index is unknown
+        class UnknownSampleTypeMappingError < ArgumentError
+          attr_reader :type
+
+          def initialize(type)
+            @type = type
+          end
+
+          def message
+            "Mapping for sample value type '#{type}' is unknown."
+          end
+        end
+      end
+    end
+  end
+end

data/lib/datadog/profiling/preload.rb

@@ -0,0 +1,4 @@
+# typed: strict
+require 'ddtrace'
+
+Datadog::Profiling.start_if_enabled

data/lib/datadog/profiling/profiler.rb

@@ -0,0 +1,33 @@
+# typed: true
+module Datadog
+  module Profiling
+    # Profiling entry point, which coordinates collectors and a scheduler
+    class Profiler
+      attr_reader \
+        :collectors,
+        :scheduler
+
+      def initialize(collectors, scheduler)
+        @collectors = collectors
+        @scheduler = scheduler
+      end
+
+      def start
+        collectors.each(&:start)
+        scheduler.start
+      end
+
+      def shutdown!
+        Datadog.logger.debug('Shutting down profiler')
+
+        collectors.each do |collector|
+          collector.enabled = false
+          collector.stop(true)
+        end
+
+        scheduler.enabled = false
+        scheduler.stop(true)
+      end
+    end
+  end
+end