Merge branch 'master' into 523-update-changelog

lib/scout_apm.rb

@@ -111,6 +111,7 @@ module ScoutApm
 require 'scout_apm/app_server_load'
 
 require 'scout_apm/ignored_uris.rb'
+require 'scout_apm/sampling.rb'
 require 'scout_apm/utils/active_record_metric_name'
 require 'scout_apm/utils/backtrace_parser'
 require 'scout_apm/utils/installed_gems'

lib/scout_apm/agent_context.rb

@@ -91,6 +91,10 @@ def logger
       @logger ||= LoggerFactory.build(config, environment)
     end
 
+    def sampling
+      @sampling ||= ScoutApm::Sampling.new(config)
+    end
+
     def ignored_uris
       @ignored_uris ||= ScoutApm::IgnoredUris.new(config.value('ignore'))
     end

lib/scout_apm/config.rb

@@ -44,9 +44,11 @@
 #                            instruments listed in this array. Default: []
 # ignore_endpoints         - An array of endpoints to ignore. These are matched as regular expressions. (supercedes 'ignore')
 # ignore_jobs              - An array of job names to ignore.
-# sample_rate              - An integer between 0 and 100. 0 means no traces are sent, 100 means all traces are sent.
+# sample_rate              - Rate to sample entire application. An integer between 0 and 100. 0 means no traces are sent, 100 means all traces are sent.
 # sample_endpoints         - An array of endpoints to sample. These are matched as regular expressions with individual sample rate of 0 to 100.
 # sample_jobs              - An array of job names with individual sample rate of 0 to 100.
+# endpoint_sample_rate     - Rate to sample all endpoints. An integer between 0 and 100. 0 means no traces are sent, 100 means all traces are sent. (supercedes 'sample_rate')
+# job_sample_rate          - Rate to sample all jobs. An integer between 0 and 100. 0 means no traces are sent, 100 means all traces are sent. (supercedes 'sample_rate')
 #
 # Any of these config settings can be set with an environment variable prefixed
 # by SCOUT_ and uppercasing the key: SCOUT_LOG_LEVEL for instance.
@@ -93,6 +95,8 @@ class Config
         'sample_rate',
         'sample_endpoints',
         'sample_jobs',
+        'endpoint_sample_rate',
+        'job_sample_rate',
         'scm_subdirectory',
         'start_resque_server_instrument',
         'ssl_cert_file',
@@ -209,6 +213,8 @@ def coerce(val)
       'sample_rate' => IntegerCoercion.new,
       'sample_endpoints' => JsonCoercion.new,
       'sample_jobs' => JsonCoercion.new,
+      'endpoint_sample_rate' => IntegerCoercion.new,
+      'job_sample_rate' => IntegerCoercion.new,
       'start_resque_server_instrument' => BooleanCoercion.new,
       'timeline_traces' => BooleanCoercion.new,
       'auto_instruments' => BooleanCoercion.new,
@@ -331,6 +337,8 @@ class ConfigDefaults
         'sample_rate'                          => 100,
         'sample_endpoints'                     => [],
         'sample_jobs'                          => [],
+        'endpoint_sample_rate'                 => 100,
+        'job_sample_rate'                      => 100,
         'start_resque_server_instrument'       => true, # still only starts if Resque is detected
         'collect_remote_ip'                    => true,
         'record_queue_time'                    => true,

lib/scout_apm/sampling.rb

@@ -0,0 +1,96 @@
+module ScoutApm
+  class Sampling
+    attr_reader :global_sample_rate, :sample_endpoints, :sample_uri_regex, :sample_jobs, :ignore_uri_regex, :ignore_jobs
+
+    def initialize(config)
+      @global_sample_rate = config.value('sample_rate')
+      # web endpoints matched prefix by regex
+      # jobs matched explicitly by name
+
+      # for now still support old config key ('ignore') for backwards compatibility
+      @ignore_endpoints = config.value('ignore').present? ? config.value('ignore') : config.value('ignore_endpoints')
+      @sample_endpoints = individual_sample_to_hash(config.value('sample_endpoints'))
+      @endpoint_sample_rate = config.value('endpoint_sample_rate')
+
+      @ignore_jobs = config.value('ignore_jobs')
+      @sample_jobs = individual_sample_to_hash(config.value('sample_jobs'))
+      @job_sample_rate = config.value('job_sample_rate')
+
+      logger.info("Sampling initialized with config: global_sample_rate: #{@global_sample_rate}, endpoint_sample_rate: #{@endpoint_sample_rate}, sample_endpoints: #{@sample_endpoints}, ignore_endpoints: #{@ignore_endpoints}, job_sample_rate: #@job_sample_rate}, sample_jobs: #{@sample_jobs}, ignore_jobs: #{@ignore_jobs}")
+    end
+
+    def drop_request?(transaction)
+      # Individual endpoint/job sampling takes precedence over ignoring.
+      # Individual endpoint/job sample rate always takes precedence over general endpoint/job rate.
+      # General endpoint/job rate always takes precedence over global sample rate
+      if transaction.job?
+        job_name = transaction.layer_finder.job.name
+        rate = job_sample_rate(job_name)
+        return sample?(rate) unless rate.nil?
+        return true if ignore_job?(job_name)
+        return sample?(@job_sample_rate) unless @job_sample_rate.nil?
+      elsif transaction.web?
+        uri = transaction.annotations[:uri]
+        rate = web_sample_rate(uri)
+        return sample?(rate) unless rate.nil?
+        return true if ignore_uri?(uri)
+        return sample?(@endpoint_sample_rate) unless @endpoint_sample_rate.nil?
+      end
+
+      # global sample check
+      if @global_sample_rate
+        return sample?(@global_sample_rate)
+      end
+
+      false # don't drop the request
+    end
+
+    def individual_sample_to_hash(sampling_config)
+      return nil if sampling_config.blank?
+      # config looks like ['/foo:50','/bar:100']. parse it into hash of string: integer
+      sample_hash = {}
+      sampling_config.each do |sample|
+        path, rate = sample.split(':')
+        sample_hash[path] = rate.to_i
+      end
+      sample_hash
+    end
+
+    def ignore_uri?(uri)
+      return false if @ignore_endpoints.blank?
+      @ignore_endpoints.each do |prefix|
+        return true if uri.start_with?(prefix)
+      end
+      false
+    end
+
+    def web_sample_rate(uri)
+      return nil if @sample_endpoints.blank?
+      @sample_endpoints.each do |prefix, rate|
+        return rate if uri.start_with?(prefix)
+      end
+      nil
+    end
+
+    def ignore_job?(job_name)
+      return false if @ignore_jobs.blank?
+      @ignore_jobs.include?(job_name)
+    end
+
+    def job_sample_rate(job_name)
+      return nil if @sample_jobs.blank?
+      @sample_jobs.fetch(job_name, nil)
+    end
+
+    def sample?(rate)
+      rand * 100 > rate
+    end
+
+    private
+
+    def logger
+      ScoutApm::Agent.instance.logger
+    end
+
+  end
+end

lib/scout_apm/tracked_request.rb

@@ -303,7 +303,11 @@ def record!
       restore_from_dump! if @agent_context.nil?
 
       # Bail out early if the user asked us to ignore this uri
-      return if @agent_context.ignored_uris.ignore?(annotations[:uri])
+      # return if @agent_context.ignored_uris.ignore?(annotations[:uri])
+      if @agent_context.sampling.drop_request?(self)
+        logger.debug("Dropping request due to sampling")
+        return
+      end
 
       apply_name_override
 

test/test_helper.rb

@@ -38,6 +38,10 @@ def value(key)
     @values[key]
   end
 
+  def values
+    @values
+  end
+
   def has_key?(key)
     @values.has_key?(key)
   end
@@ -169,3 +173,28 @@ def assert_false(thing)
 class Minitest::Test
   include CustomAsserts
 end
+
+class FakeTrackedRequest
+  def self.new_web_request(uri)
+    context = ScoutApm::Agent.instance.context
+    fake_store = ScoutApm::FakeStore.new
+    req = ScoutApm::TrackedRequest.new(context, fake_store)
+
+    first_layer = ScoutApm::Layer.new("Controller", "index")
+    req.start_layer(first_layer)
+    req.annotate_request(:uri => uri)
+
+    req
+  end
+
+  def self.new_job_request(job_name)
+    context = ScoutApm::Agent.instance.context
+    fake_store = ScoutApm::FakeStore.new
+    req = ScoutApm::TrackedRequest.new(context, fake_store)
+
+    first_layer = ScoutApm::Layer.new("Job", job_name)
+    req.start_layer(first_layer)
+
+    req
+  end
+end