diff options
Diffstat (limited to 'app/presenters/ci/build_runner_presenter.rb')
-rw-r--r-- | app/presenters/ci/build_runner_presenter.rb | 18 |
1 files changed, 3 insertions, 15 deletions
diff --git a/app/presenters/ci/build_runner_presenter.rb b/app/presenters/ci/build_runner_presenter.rb index ffa33dc9f15..dbb77143e2e 100644 --- a/app/presenters/ci/build_runner_presenter.rb +++ b/app/presenters/ci/build_runner_presenter.rb @@ -93,22 +93,10 @@ module Ci end def refspec_for_persistent_ref - # - # End-to-end test coverage for CI fetching seems to not be strong, so we - # are using a feature flag here to close the confidence gap. My (JV) - # confidence about the change is very high but if something is wrong - # with it after all, this would cause all CI jobs on gitlab.com to fail. - # - # The roll-out will be tracked in + # Use persistent_ref.sha because it sometimes causes 'git fetch' to do + # less work. See # https://gitlab.com/gitlab-com/gl-infra/scalability/-/issues/746. - # - if Feature.enabled?(:scalability_ci_fetch_sha, type: :ops) - # Use persistent_ref.sha because it causes 'git fetch' to do less work. - # See https://gitlab.com/gitlab-com/gl-infra/scalability/-/issues/746. - "+#{pipeline.persistent_ref.sha}:#{pipeline.persistent_ref.path}" - else - "+#{pipeline.persistent_ref.path}:#{pipeline.persistent_ref.path}" - end + "+#{pipeline.persistent_ref.sha}:#{pipeline.persistent_ref.path}" end def persistent_ref_exist? |