Remove redundant pipeline stages from the database

db/post_migrate/20180119121225_remove_redundant_pipeline_stages.rb

@@ -0,0 +1,26 @@
+class RemoveRedundantPipelineStages < ActiveRecord::Migration
+  include Gitlab::Database::MigrationHelpers
+
+  DOWNTIME = false
+
+  def up
+   redundant_stages_ids = <<~SQL
+      SELECT id FROM ci_stages a WHERE (
+        SELECT COUNT(*) FROM ci_stages b
+          WHERE a.pipeline_id = b.pipeline_id AND a.name = b.name
+      ) > 1
+    SQL
+
+    execute <<~SQL
+      UPDATE ci_builds SET stage_id = NULL WHERE ci_builds.stage_id IN (#{redundant_stages_ids})
+    SQL
+
+    execute <<~SQL
+      DELETE FROM ci_stages WHERE ci_stages.id IN (#{redundant_stages_ids})
+    SQL
+  end
+
+  def down
+    # noop
+  end
+end

db/schema.rb

@@ -11,7 +11,7 @@
 #
 # It's strongly recommended that you check this file into your version control system.
 
-ActiveRecord::Schema.define(version: 20180113220114) do
+ActiveRecord::Schema.define(version: 20180119121225) do
 
   # These are extensions that must be enabled in order to support this database
   enable_extension "plpgsql"

spec/migrations/remove_redundant_pipeline_stages_spec.rb

@@ -0,0 +1,39 @@
+require 'spec_helper'
+require Rails.root.join('db', 'post_migrate', '20180119121225_remove_redundant_pipeline_stages.rb')
+
+describe RemoveRedundantPipelineStages, :migration do
+  let(:projects) { table(:projects) }
+  let(:pipelines) { table(:ci_pipelines) }
+  let(:stages) { table(:ci_stages) }
+  let(:builds) { table(:ci_builds) }
+
+  before do
+    projects.create!(id: 123, name: 'gitlab', path: 'gitlab-ce')
+    pipelines.create!(id: 234, project_id: 123, ref: 'master', sha: 'adf43c3a')
+
+    stages.create!(id: 6, project_id: 123, pipeline_id: 234, name: 'build')
+    stages.create!(id: 10, project_id: 123, pipeline_id: 234, name: 'build')
+    stages.create!(id: 21, project_id: 123, pipeline_id: 234, name: 'build')
+    stages.create!(id: 41, project_id: 123, pipeline_id: 234, name: 'test')
+    stages.create!(id: 62, project_id: 123, pipeline_id: 234, name: 'test')
+    stages.create!(id: 102, project_id: 123, pipeline_id: 234, name: 'deploy')
+
+    builds.create!(id: 1, commit_id: 234, project_id: 123, stage_id: 10)
+    builds.create!(id: 2, commit_id: 234, project_id: 123, stage_id: 21)
+    builds.create!(id: 3, commit_id: 234, project_id: 123, stage_id: 21)
+    builds.create!(id: 4, commit_id: 234, project_id: 123, stage_id: 41)
+    builds.create!(id: 5, commit_id: 234, project_id: 123, stage_id: 62)
+    builds.create!(id: 6, commit_id: 234, project_id: 123, stage_id: 102)
+  end
+
+  it 'removes ambiguous stages and preserves builds' do
+    expect(stages.all.count).to eq 6
+    expect(builds.all.count).to eq 6
+
+    migrate!
+
+    expect(stages.all.count).to eq 1
+    expect(builds.all.count).to eq 6
+    expect(builds.all.pluck(:stage_id).compact).to eq [102]
+  end
+end