Load project ids from tracking database when searching for batch

Closes gitlab-org/gitlab-ee#5564

app/workers/repository_check/batch_worker.rb

 module RepositoryCheck
   class BatchWorker
+    prepend ::EE::RepositoryCheck::BatchWorker
+
     include ApplicationWorker
     include CronjobQueue
 

ee/app/workers/ee/repository_check/batch_worker.rb

+module EE
+  module RepositoryCheck
+    module BatchWorker
+      extend ActiveSupport::Concern
+      extend ::Gitlab::Utils::Override
+
+      private
+
+      override :never_checked_project_ids
+      def never_checked_project_ids(batch_size)
+        return super unless ::Gitlab::Geo.secondary?
+
+        Geo::ProjectRegistry.synced_repos.synced_wikis
+          .where(last_repository_check_at: nil)
+          .where('last_repository_synced_at < ?', 24.hours.ago)
+          .where('last_wiki_synced_at < ?', 24.hours.ago)
+          .limit(batch_size).pluck(:project_id)
+      end
+
+      override :old_checked_project_ids
+      def old_checked_project_ids(batch_size)
+        return super unless ::Gitlab::Geo.secondary?
+
+        Geo::ProjectRegistry.synced_repos.synced_wikis
+          .where('last_repository_check_at < ?', 1.month.ago)
+          .reorder(last_repository_check_at: :asc)
+          .limit(batch_size).pluck(:project_id)
+      end
+    end
+  end
+end

ee/changelogs/unreleased/tc-repo-verify-mails-ee.yml

+---
+title: Run repository verification on Geo secondary
+merge_request: 5550
+author:
+type: added

ee/spec/workers/ee/repository_check/batch_worker_spec.rb

+require 'spec_helper'
+
+describe EE::RepositoryCheck::BatchWorker do
+  include ::EE::GeoHelpers
+
+  subject(:worker) { RepositoryCheck::BatchWorker.new }
+
+  context 'Geo primary' do
+    set(:primary) { create(:geo_node, :primary) }
+
+    before do
+      stub_current_geo_node(primary)
+    end
+
+    it 'loads project ids from main database' do
+      projects = create_list(:project, 3, created_at: 1.week.ago)
+
+      expect(worker.perform).to eq(projects.map(&:id))
+    end
+  end
+
+  context 'Geo secondary' do
+    set(:secondary) { create(:geo_node) }
+
+    before do
+      stub_current_geo_node(secondary)
+    end
+
+    it 'loads project ids from tracking database' do
+      project_registries = create_list(:geo_project_registry, 3, :synced)
+
+      expect(worker.perform).to eq(project_registries.map(&:project_id))
+    end
+
+    it 'loads project ids that were checked more than a month ago from tracking database' do
+      project_registries = create_list(:geo_project_registry, 3, :synced,
+                                       last_repository_check_failed: false,
+                                       last_repository_check_at: 42.days.ago)
+
+      expect(worker.perform).to eq(project_registries.map(&:project_id))
+    end
+  end
+end