[FF] track_ai_pipeline_results_viewed -- view_ai_pipeline_results internal event tracking

Summary

Roll out the feature currently behind the track_ai_pipeline_results_viewed feature flag.

This flag gates enqueuing Ci::TrackAiPipelineResultsViewedWorker from the pipeline show page. The worker fires the view_ai_pipeline_results internal event (the "Pipeline results reviewed" funnel metric) the first time a user views the results of a pipeline whose CI config was AI-generated.

  • DRI: @sasharma13
  • Team Slack channel: #g_pipeline-authoring

What could go wrong?

Low blast radius. The flag only controls whether a low-urgency, idempotent tracking worker is enqueued from the pipeline page; no user-facing behaviour changes and no data is destroyed. The flag exists mainly to de-risk introducing a new worker class across the web/Sidekiq deploy window (avoids NameError: uninitialized constant if Sidekiq nodes lag web nodes). Worst case if misbehaving: the metric is over- or under-counted, mitigated by disabling the flag.

Rollout

Run all production /chatops in #production and cross-post to #g_pipeline-authoring.

Non-production

/chatops gitlab run feature set track_ai_pipeline_results_viewed true --dev --pre --staging --staging-ref

Production (percentage rollout on the project actor, wait >=15 min between steps):

/chatops gitlab run feature set track_ai_pipeline_results_viewed <percentage> --actors

Cleanup

Remove the flag once deemed stable: delete the flag and its YAML definition, then:

/chatops gitlab run release check <merge-request-url> <milestone>
/chatops gitlab run feature delete track_ai_pipeline_results_viewed --dev --pre --staging --staging-ref --production

Rollback

/chatops gitlab run feature set track_ai_pipeline_results_viewed false
/chatops gitlab run feature set track_ai_pipeline_results_viewed false --dev --pre --staging --staging-ref
Edited by Sahil Sharma