ddtrace 0.40.0 → 0.54.1

data/lib/ddtrace/span.rb

@@ -1,13 +1,18 @@
-require 'time'
-require 'thread'
+# typed: true
+# frozen_string_literal: true
 
+require 'time'
 require 'ddtrace/utils'
+require 'ddtrace/ext/distributed'
+require 'ddtrace/ext/environment'
 require 'ddtrace/ext/errors'
+require 'ddtrace/ext/http'
+require 'ddtrace/ext/net'
 require 'ddtrace/ext/priority'
-require 'ddtrace/environment'
 require 'ddtrace/analytics'
 require 'ddtrace/forced_tracing'
 require 'ddtrace/diagnostics/health'
+require 'ddtrace/utils/time'
 
 module Datadog
   # Represents a logical unit of work in the system. Each trace consists of one or more spans.
@@ -24,29 +29,45 @@ module Datadog
     # The max value for a \Span identifier.
     # Span and trace identifiers should be strictly positive and strictly inferior to this limit.
     #
-    # Limited to 63-bit positive integers, as some other languages might be limited to this,
-    # and IDs need to be easy to port across various languages and platforms.
-    MAX_ID = 2**63
+    # Limited to +2<<62-1+ positive integers, as Ruby is able to represent such numbers "inline",
+    # inside a +VALUE+ scalar, thus not requiring memory allocation.
+    #
+    # The range of IDs also has to consider portability across different languages and platforms.
+    RUBY_MAX_ID = (1 << 62) - 1
 
     # While we only generate 63-bit integers due to limitations in other languages, we support
     # parsing 64-bit integers for distributed tracing since an upstream system may generate one
-    EXTERNAL_MAX_ID = 2**64
+    EXTERNAL_MAX_ID = 1 << 64
 
     # This limit is for numeric tags because uint64 could end up rounded.
-    NUMERIC_TAG_SIZE_RANGE = (-2**53..2**53)
+    NUMERIC_TAG_SIZE_RANGE = (-1 << 53..1 << 53).freeze
+
+    # Some associated values should always be sent as Tags, never as Metrics, regardless
+    # if their value is numeric or not.
+    # The Datadog agent will look for these values only as Tags, not Metrics.
+    # @see https://github.com/DataDog/datadog-agent/blob/2ae2cdd315bcda53166dd8fa0dedcfc448087b9d/pkg/trace/stats/aggregation.go#L13-L17
+    ENSURE_AGENT_TAGS = {
+      Ext::DistributedTracing::ORIGIN_KEY => true,
+      Ext::Environment::TAG_VERSION => true,
+      Ext::HTTP::STATUS_CODE => true,
+      Ext::NET::TAG_HOSTNAME => true
+    }.freeze
 
     attr_accessor :name, :service, :resource, :span_type,
-                  :start_time, :end_time,
                   :span_id, :trace_id, :parent_id,
                   :status, :sampled,
                   :tracer, :context
 
-    attr_reader :parent
+    attr_reader :parent, :start_time, :end_time, :resource_container
+
+    attr_writer :duration
+
     # Create a new span linked to the given tracer. Call the \Tracer method <tt>start_span()</tt>
     # and then <tt>finish()</tt> once the tracer operation is over.
     #
     # * +service+: the service name for this span
-    # * +resource+: the resource this span refers, or +name+ if it's missing
+    # * +resource+: the resource this span refers, or +name+ if it's missing.
+    #     +nil+ can be used as a placeholder, when the resource value is not yet known at +#initialize+ time.
     # * +span_type+: the type of the span (such as +http+, +db+ and so on)
     # * +parent_id+: the identifier of the parent span
     # * +trace_id+: the identifier of the root span for this trace
@@ -72,11 +93,20 @@ module Datadog
       @parent = nil
       @sampled = true
 
-      @start_time = nil # set by Tracer.start_span
-      @end_time = nil # set by Span.finish
-
       @allocation_count_start = now_allocations
       @allocation_count_finish = @allocation_count_start
+
+      # start_time and end_time track wall clock. In Ruby, wall clock
+      # has less accuracy than monotonic clock, so if possible we look to only use wall clock
+      # to measure duration when a time is supplied by the user, or if monotonic clock
+      # is unsupported.
+      @start_time = nil
+      @end_time = nil
+
+      # duration_start and duration_end track monotonic clock, and may remain nil in cases where it
+      # is known that we have to use wall clock to measure duration.
+      @duration_start = nil
+      @duration_end = nil
     end
 
     # Set the given key / value tag pair on the span. Keys and values
@@ -87,11 +117,8 @@ module Datadog
       # Keys must be unique between tags and metrics
       @metrics.delete(key)
 
-      # Ensure `http.status_code` is always a string so it is added to
-      #   @meta instead of @metrics
-      # DEV: This is necessary because the agent looks to `meta['http.status_code']` for
-      #   tagging necessary metrics
-      value = value.to_s if key == Ext::HTTP::STATUS_CODE
+      # DEV: This is necessary because the agent looks at `meta[key]`, not `metrics[key]`.
+      value = value.to_s if ENSURE_AGENT_TAGS[key]
 
       # NOTE: Adding numeric tags as metrics is stop-gap support
       #       for numeric typed tags. Eventually they will become
@@ -160,6 +187,28 @@ module Datadog
       set_tag(Ext::Errors::STACK, e.backtrace) unless e.backtrace.empty?
     end
 
+    # Mark the span started at the current time.
+    def start(start_time = nil)
+      # A span should not be started twice. However, this is existing
+      # behavior and so we maintain it for backward compatibility for those
+      # who are using async manual instrumentation that may rely on this
+
+      @start_time = start_time || Utils::Time.now.utc
+      @duration_start = start_time.nil? ? duration_marker : nil
+
+      self
+    end
+
+    # for backwards compatibility
+    def start_time=(time)
+      time.tap { start(time) }
+    end
+
+    # for backwards compatibility
+    def end_time=(time)
+      time.tap { finish(time) }
+    end
+
     # Mark the span finished at the current time and submit it.
     def finish(finish_time = nil)
       # A span should not be finished twice. Note that this is not thread-safe,
@@ -170,12 +219,15 @@ module Datadog
 
       @allocation_count_finish = now_allocations
 
-      # Provide a default start_time if unset, but this should have been set by start_span.
-      # Using now here causes 0-duration spans, still, this is expected, as we never
-      # explicitely say when it started.
-      @start_time ||= Time.now.utc
+      now = Utils::Time.now.utc
 
-      @end_time = finish_time.nil? ? Time.now.utc : finish_time # finish this
+      # Provide a default start_time if unset.
+      # Using `now` here causes duration to be 0; this is expected
+      # behavior when start_time is unknown.
+      start(finish_time || now) unless started?
+
+      @end_time = finish_time || now
+      @duration_end = finish_time.nil? ? duration_marker : nil
 
       # Finish does not really do anything if the span is not bound to a tracer and a context.
       return self if @tracer.nil? || @context.nil?
@@ -195,11 +247,6 @@ module Datadog
       self
     end
 
-    # Return whether the span is finished or not.
-    def finished?
-      !@end_time.nil?
-    end
-
     # Return a string representation of the span.
     def to_s
       "Span(name:#{@name},sid:#{@span_id},tid:#{@trace_id},pid:#{@parent_id})"
@@ -246,19 +293,76 @@ module Datadog
         error: @status
       }
 
-      if !@start_time.nil? && !@end_time.nil?
-        h[:start] = (@start_time.to_f * 1e9).to_i
-        h[:duration] = ((@end_time - @start_time) * 1e9).to_i
+      if finished?
+        h[:start] = start_time_nano
+        h[:duration] = duration_nano
       end
 
       h
     end
 
+    # MessagePack serializer interface. Making this object
+    # respond to `#to_msgpack` allows it to be automatically
+    # serialized by MessagePack.
+    #
+    # This is more efficient than doing +MessagePack.pack(span.to_hash)+
+    # as we don't have to create an intermediate Hash.
+    #
+    # @param packer [MessagePack::Packer] serialization buffer, can be +nil+ with JRuby
+    def to_msgpack(packer = nil)
+      # As of 1.3.3, JRuby implementation doesn't pass an existing packer
+      packer ||= MessagePack::Packer.new
+
+      if finished?
+        packer.write_map_header(13) # Set header with how many elements in the map
+
+        packer.write('start')
+        packer.write(start_time_nano)
+
+        packer.write('duration')
+        packer.write(duration_nano)
+      else
+        packer.write_map_header(11) # Set header with how many elements in the map
+      end
+
+      # DEV: We use strings as keys here, instead of symbols, as
+      # DEV: MessagePack will ultimately convert them to strings.
+      # DEV: By providing strings directly, we skip this indirection operation.
+      packer.write('span_id')
+      packer.write(@span_id)
+      packer.write('parent_id')
+      packer.write(@parent_id)
+      packer.write('trace_id')
+      packer.write(@trace_id)
+      packer.write('name')
+      packer.write(@name)
+      packer.write('service')
+      packer.write(@service)
+      packer.write('resource')
+      packer.write(@resource)
+      packer.write('type')
+      packer.write(@span_type)
+      packer.write('meta')
+      packer.write(@meta)
+      packer.write('metrics')
+      packer.write(@metrics)
+      packer.write('allocations')
+      packer.write(allocations)
+      packer.write('error')
+      packer.write(@status)
+      packer
+    end
+
+    # JSON serializer interface.
+    # Used by older version of the transport.
+    def to_json(*args)
+      to_hash.to_json(*args)
+    end
+
     # Return a human readable version of the span
     def pretty_print(q)
-      start_time = (@start_time.to_f * 1e9).to_i rescue '-'
-      end_time = (@end_time.to_f * 1e9).to_i rescue '-'
-      duration = ((@end_time - @start_time) * 1e9).to_i rescue 0
+      start_time = (self.start_time.to_f * 1e9).to_i
+      end_time = (self.end_time.to_f * 1e9).to_i
       q.group 0 do
         q.breakable
         q.text "Name: #{@name}\n"
@@ -271,7 +375,7 @@ module Datadog
         q.text "Error: #{@status}\n"
         q.text "Start: #{start_time}\n"
         q.text "End: #{end_time}\n"
-        q.text "Duration: #{duration}\n"
+        q.text "Duration: #{duration.to_f if finished?}\n"
         q.text "Allocations: #{allocations}\n"
         q.group(2, 'Tags: [', "]\n") do
           q.breakable
@@ -288,8 +392,30 @@ module Datadog
       end
     end
 
+    # Return whether the duration is started or not
+    def started?
+      !@start_time.nil?
+    end
+
+    # Return whether the duration is finished or not.
+    def finished?
+      !@end_time.nil?
+    end
+
+    def duration
+      if @duration_end.nil? || @duration_start.nil?
+        @end_time - @start_time
+      else
+        @duration_end - @duration_start
+      end
+    end
+
     private
 
+    def duration_marker
+      Utils::Time.get_time
+    end
+
     if defined?(JRUBY_VERSION) || Gem::Version.new(RUBY_VERSION) < Gem::Version.new(VERSION::MINIMUM_RUBY_VERSION)
       def now_allocations
         0
@@ -303,5 +429,17 @@ module Datadog
         GC.stat(:total_allocated_objects)
       end
     end
+
+    # Used for serialization
+    # @return [Integer] in nanoseconds since Epoch
+    def start_time_nano
+      @start_time.to_i * 1000000000 + @start_time.nsec
+    end
+
+    # Used for serialization
+    # @return [Integer] in nanoseconds since Epoch
+    def duration_nano
+      (duration * 1e9).to_i
+    end
   end
 end

data/lib/ddtrace/sync_writer.rb

@@ -1,11 +1,19 @@
+# typed: true
 require 'ddtrace/ext/net'
-require 'ddtrace/runtime/socket'
+require 'datadog/core/environment/socket'
 require 'ddtrace/runtime/metrics'
+require 'ddtrace/utils/only_once'
 
 module Datadog
   # SyncWriter flushes both services and traces synchronously
   # DEV: To be replaced by Datadog::Workers::TraceWriter.
+  #
+  # Note: If you're wondering if this class is used at all, since there are no other references to it on the codebase,
+  # the separate `datadog-lambda` uses it as of February 2021:
+  # <https://github.com/DataDog/datadog-lambda-rb/blob/c15f0f0916c90123416dc44e7d6800ef4a7cfdbf/lib/datadog/lambda.rb#L38>
   class SyncWriter
+    DEPRECATION_WARN_ONLY_ONCE = Datadog::Utils::OnlyOnce.new
+
     attr_reader \
       :priority_sampler,
       :transport
@@ -13,7 +21,8 @@ module Datadog
     def initialize(options = {})
       @transport = options.fetch(:transport) do
         transport_options = options.fetch(:transport_options, {})
-        Transport::HTTP.default(transport_options)
+        transport_options[:agent_settings] = options[:agent_settings] if options.key?(:agent_settings)
+        Transport::HTTP.default(**transport_options)
       end
 
       @priority_sampler = options.fetch(:priority_sampler, nil)
@@ -21,17 +30,15 @@ module Datadog
 
     def write(trace, services = nil)
       unless services.nil?
-        Datadog::Patcher.do_once('SyncWriter#write') do
+        DEPRECATION_WARN_ONLY_ONCE.run do
           Datadog.logger.warn(%(
             write: Writing services has been deprecated and no longer need to be provided.
-            write(traces, services) can be updted to write(traces)
+            write(traces, services) can be updated to write(traces)
           ))
         end
       end
 
-      perform_concurrently(
-        proc { flush_trace(trace) }
-      )
+      flush_trace(trace)
     rescue => e
       Datadog.logger.debug(e)
     end
@@ -44,23 +51,18 @@ module Datadog
 
     private
 
-    def perform_concurrently(*tasks)
-      tasks.map { |task| Thread.new(&task) }.each(&:join)
-    end
-
     def flush_trace(trace)
       processed_traces = Pipeline.process!([trace])
       return if processed_traces.empty?
+
       inject_hostname!(processed_traces.first) if Datadog.configuration.report_hostname
       transport.send_traces(processed_traces)
     end
 
     def inject_hostname!(trace)
       unless trace.first.nil?
-        hostname = Datadog::Runtime::Socket.hostname
-        unless hostname.nil? || hostname.empty?
-          trace.first.set_tag(Ext::NET::TAG_HOSTNAME, hostname)
-        end
+        hostname = Datadog::Core::Environment::Socket.hostname
+        trace.first.set_tag(Ext::NET::TAG_HOSTNAME, hostname) unless hostname.nil? || hostname.empty?
       end
     end
   end

data/lib/ddtrace/tasks/exec.rb

@@ -0,0 +1,47 @@
+# typed: true
+module Datadog
+  module Tasks
+    # Wraps command with Datadog tracing
+    class Exec
+      attr_reader :args
+
+      def initialize(args)
+        @args = args
+      end
+
+      def run
+        set_rubyopt!
+        exec_with_error_handling(args)
+      end
+
+      def rubyopts
+        [
+          '-rddtrace/profiling/preload'
+        ]
+      end
+
+      private
+
+      def set_rubyopt!
+        existing_rubyopt = ENV['RUBYOPT']
+
+        ENV['RUBYOPT'] = existing_rubyopt ? "#{existing_rubyopt} #{rubyopts.join(' ')}" : rubyopts.join(' ')
+      end
+
+      # If there's an error here, rather than throwing a cryptic stack trace, let's instead have clearer messages, and
+      # follow the same status codes as the shell uses
+      # See also:
+      # * https://www.gnu.org/software/bash/manual/html_node/Exit-Status.html
+      # * https://github.com/rubygems/rubygems/blob/dd93966cac224532035deda533cba2685dfa30cc/bundler/lib/bundler/cli/exec.rb#L45
+      def exec_with_error_handling(args)
+        Kernel.exec(*args)
+      rescue Errno::ENOENT => e
+        Kernel.warn "ddtracerb exec failed: #{e.message} (command was '#{args.join(' ')}')"
+        Kernel.exit 127
+      rescue Errno::EACCES, Errno::ENOEXEC => e
+        Kernel.warn "ddtracerb exec failed: #{e.message} (command was '#{args.join(' ')}')"
+        Kernel.exit 126
+      end
+    end
+  end
+end

data/lib/ddtrace/tasks/help.rb

@@ -0,0 +1,15 @@
+# typed: true
+module Datadog
+  module Tasks
+    # Prints help message for usage of `ddtrace`
+    class Help
+      def run
+        puts %(
+Usage: ddtrace [command] [arguments]
+  exec [command]: Executes command with tracing & profiling preloaded.
+  help:           Prints this help message.
+        )
+      end
+    end
+  end
+end