Merge branch 'bvl-remove-apdex-counters-ff' into 'master'

Remove the rails_request_apdex_counters feature flag

See merge request gitlab-org/gitlab!72427

config/feature_flags/development/request_apdex_counters.yml

----
-name: request_apdex_counters
-introduced_by_url: https://gitlab.com/gitlab-org/gitlab/-/merge_requests/69154
-rollout_issue_url: https://gitlab.com/gitlab-com/gl-infra/scalability/-/issues/1099
-milestone: '14.3'
-type: development
-group: team::Scalability
-default_enabled: false

doc/administration/monitoring/prometheus/gitlab_metrics.md

@@ -276,6 +276,8 @@ configuration option in `gitlab.yml`. These metrics are served from the
 | `geo_uploads_synced`               | Gauge   | 14.1  | Number of uploads synced on secondary | `url` |
 | `geo_uploads_failed`               | Gauge   | 14.1  | Number of syncable uploads failed to sync on secondary | `url` |
 | `geo_uploads_registry`             | Gauge   | 14.1  | Number of uploads in the registry | `url` |
+| `gitlab_sli:rails_request_apdex:total` | Counter | 14.4 | The number of request-apdex measurements, [more information the development documentation](../../../development/application_slis/rails_request_apdex.md) | `endpoint_id`, `feature_category`, `request_urgency` |
+| `gitlab_sli:rails_request_apdex:success_total` | Counter | 14.4 | The number of succesful requests that met the target duration for their urgency. Devide by `gitlab_sli:rails_requests_apdex:total` to get a success ratio | `endpoint_id`, `feature_category`, `request_urgency` |
 
 ## Database load balancing metrics **(PREMIUM SELF)**
 

lib/gitlab/metrics/rails_slis.rb

@@ -4,12 +4,7 @@ module Gitlab
   module Metrics
     module RailsSlis
       class << self
-        def request_apdex_counters_enabled?
-          Feature.enabled?(:request_apdex_counters)
-        end
-
         def initialize_request_slis_if_needed!
-          return unless request_apdex_counters_enabled?
           return if Gitlab::Metrics::Sli.initialized?(:rails_request_apdex)
 
           Gitlab::Metrics::Sli.initialize_sli(:rails_request_apdex, possible_request_labels)

lib/gitlab/metrics/requests_rack_middleware.rb

@@ -79,7 +79,7 @@ module Gitlab
           if !health_endpoint && ::Gitlab::Metrics.record_duration_for_status?(status)
             self.class.http_request_duration_seconds.observe({ method: method }, elapsed)
 
-            record_apdex_if_needed(env, elapsed)
+            record_apdex(env, elapsed)
           end
 
           [status, headers, body]
@@ -113,9 +113,7 @@ module Gitlab
         ::Gitlab::ApplicationContext.current_context_attribute(:caller_id)
       end
 
-      def record_apdex_if_needed(env, elapsed)
-        return unless Gitlab::Metrics::RailsSlis.request_apdex_counters_enabled?
-
+      def record_apdex(env, elapsed)
         urgency = urgency_for_env(env)
 
         Gitlab::Metrics::RailsSlis.request_apdex.increment(

spec/lib/gitlab/metrics/rails_slis_spec.rb

@@ -39,15 +39,6 @@ RSpec.describe Gitlab::Metrics::RailsSlis do
 
       described_class.initialize_request_slis_if_needed!
     end
-
-    it 'does not initialize anything if the feature flag is disabled' do
-      stub_feature_flags(request_apdex_counters: false)
-
-      expect(Gitlab::Metrics::Sli).not_to receive(:initialize_sli)
-      expect(Gitlab::Metrics::Sli).not_to receive(:initialized?)
-
-      described_class.initialize_request_slis_if_needed!
-    end
   end
 
   describe '.request_apdex' do