Merge branch '522011-backfill-finish-onboarding-saml' into 'master'

Remove existing users from onboarding if they have group saml

See merge request https://gitlab.com/gitlab-org/gitlab/-/merge_requests/183559



Merged-by: Tiger Watson <twatson@gitlab.com>
Approved-by: Surabhi Suman <ssuman@gitlab.com>
Approved-by: Imre Farkas <ifarkas@gitlab.com>
Approved-by: Tetiana Chupryna <tchupryna@gitlab.com>
Approved-by: Tiger Watson <twatson@gitlab.com>
Reviewed-by: Doug Stull <dstull@gitlab.com>
Reviewed-by: Surabhi Suman <ssuman@gitlab.com>
Co-authored-by: Doug Stull <dstull@gitlab.com>

db/docs/batched_background_migrations/backfill_finish_onboarding_for_group_saml.yml

+---
+migration_job_name: BackfillFinishOnboardingForGroupSaml
+description: Finishes the onboarding for a user if they are now attached to group saml
+feature_category: onboarding
+introduced_by_url: https://gitlab.com/gitlab-org/gitlab/-/merge_requests/183559
+milestone: '17.10'
+queued_migration_version: 20250305135537
+finalized_by: # version of the migration that finalized this BBM

db/post_migrate/20250305135537_queue_backfill_finish_onboarding_for_group_saml.rb

+# frozen_string_literal: true
+
+class QueueBackfillFinishOnboardingForGroupSaml < Gitlab::Database::Migration[2.2]
+  milestone '17.10'
+
+  restrict_gitlab_migration gitlab_schema: :gitlab_main
+
+  MIGRATION = "BackfillFinishOnboardingForGroupSaml"
+  DELAY_INTERVAL = 2.minutes
+  BATCH_SIZE = 3_000
+  SUB_BATCH_SIZE = 250
+  MAX_BATCH_SIZE = 10_000
+
+  def up
+    queue_batched_background_migration(
+      MIGRATION,
+      :identities,
+      :id,
+      job_interval: DELAY_INTERVAL,
+      batch_size: BATCH_SIZE,
+      sub_batch_size: SUB_BATCH_SIZE,
+      max_batch_size: MAX_BATCH_SIZE
+    )
+  end
+
+  def down
+    delete_batched_background_migration(MIGRATION, :identities, :id, [])
+  end
+end

db/schema_migrations/20250305135537

+ad139258454e70107e8019c35171046e9b788ae272d34c1c68d182d5e1a43006
\ No newline at end of file

lib/gitlab/background_migration/backfill_finish_onboarding_for_group_saml.rb

+# frozen_string_literal: true
+
+module Gitlab
+  module BackgroundMigration
+    class BackfillFinishOnboardingForGroupSaml < BatchedMigrationJob
+      operation_name :backfill_finish_onboarding_for_group_saml # This is used as the key on collecting metrics
+      scope_to ->(relation) { relation.where(provider: 'group_saml') }
+      feature_category :onboarding
+
+      class User < ApplicationRecord
+        self.table_name = :users
+      end
+
+      def perform
+        each_sub_batch do |sub_batch|
+          # First get the user IDs from the identities query
+          user_ids = sub_batch
+                       .joins("INNER JOIN users ON users.id = identities.user_id")
+                       .where(provider: 'group_saml', users: { onboarding_in_progress: true })
+                       .pluck(:user_id)
+
+          # Then update those users
+          User.where(id: user_ids).update_all(onboarding_in_progress: false)
+        end
+      end
+    end
+  end
+end

spec/lib/gitlab/background_migration/backfill_finish_onboarding_for_group_saml_spec.rb

+# frozen_string_literal: true
+
+require 'spec_helper'
+
+RSpec.describe Gitlab::BackgroundMigration::BackfillFinishOnboardingForGroupSaml, feature_category: :onboarding do
+  let(:users) { table(:users) }
+  let(:identities) { table(:identities) }
+
+  let(:first_user_with_saml) do
+    users.create!(email: 'user1@example.com', projects_limit: 0, onboarding_in_progress: true)
+  end
+
+  let!(:first_user_identity) { identities.create!(user_id: first_user_with_saml.id, provider: 'group_saml') }
+
+  let!(:user_with_saml_not_in_onboarding) do
+    record = users.create!(email: 'user2@example.com', projects_limit: 0, onboarding_in_progress: false)
+    identities.create!(user_id: record.id, provider: 'group_saml')
+    record
+  end
+
+  let!(:user_with_identity_not_group_saml) do
+    record = users.create!(email: 'user3@example.com', projects_limit: 0, onboarding_in_progress: true)
+    identities.create!(user_id: record.id, provider: 'foo')
+    record
+  end
+
+  let!(:last_user_no_identity) do
+    users.create!(email: 'user4@example.com', projects_limit: 0, onboarding_in_progress: true)
+  end
+
+  let(:last_user_with_identity) do
+    users.create!(email: 'user5@example.com', projects_limit: 0, onboarding_in_progress: true)
+  end
+
+  let!(:last_user_not_saml_identity) { identities.create!(user_id: last_user_with_identity.id, provider: 'foo') }
+  let!(:last_user_identity) { identities.create!(user_id: last_user_with_identity.id, provider: 'group_saml') }
+
+  subject(:migration) do
+    described_class.new(
+      start_id: first_user_identity.id,
+      end_id: last_user_identity.id,
+      batch_table: :identities,
+      batch_column: :id,
+      sub_batch_size: 100,
+      pause_ms: 0,
+      connection: ApplicationRecord.connection
+    )
+  end
+
+  describe '#perform' do
+    it 'updates the correct data' do
+      migration.perform
+
+      expect(first_user_with_saml.reload.onboarding_in_progress).to be(false)
+      expect(user_with_saml_not_in_onboarding.reload.onboarding_in_progress).to be(false)
+      expect(user_with_identity_not_group_saml.reload.onboarding_in_progress).to be(true)
+      expect(last_user_no_identity.reload.onboarding_in_progress).to be(true)
+      expect(last_user_with_identity.reload.onboarding_in_progress).to be(false)
+    end
+  end
+end

spec/migrations/20250305135537_queue_backfill_finish_onboarding_for_group_saml_spec.rb

+# frozen_string_literal: true
+
+require 'spec_helper'
+require_migration!
+
+RSpec.describe QueueBackfillFinishOnboardingForGroupSaml, migration: :gitlab_main_cell, feature_category: :onboarding do
+  let!(:batched_migration) { described_class::MIGRATION }
+
+  it 'schedules a new batched migration' do
+    reversible_migration do |migration|
+      migration.before -> {
+        expect(batched_migration).not_to have_scheduled_batched_migration
+      }
+
+      attributes = {
+        table_name: :identities,
+        column_name: :id,
+        interval: described_class::DELAY_INTERVAL,
+        batch_size: described_class::BATCH_SIZE,
+        sub_batch_size: described_class::SUB_BATCH_SIZE,
+        max_batch_size: described_class::MAX_BATCH_SIZE
+      }
+
+      migration.after -> {
+        expect(batched_migration).to have_scheduled_batched_migration(attributes)
+      }
+    end
+  end
+end