Merge branch 'add-background-migrations-for-not-archived-traces' into 'master'

Add background migrations to archive legacy job traces

Closes #46642

See merge request gitlab-org/gitlab-ce!19194

app/models/ci/build.rb

@@ -55,6 +55,11 @@ module Ci
       where('(artifacts_file IS NOT NULL AND artifacts_file <> ?) OR EXISTS (?)',
         '', Ci::JobArtifact.select(1).where('ci_builds.id = ci_job_artifacts.job_id').archive)
     end
+
+    scope :without_archived_trace, ->() do
+      where('NOT EXISTS (?)', Ci::JobArtifact.select(1).where('ci_builds.id = ci_job_artifacts.job_id').trace)
+    end
+
     scope :with_artifacts_stored_locally, -> { with_artifacts_archive.where(artifacts_file_store: [nil, LegacyArtifactUploader::Store::LOCAL]) }
     scope :with_artifacts_not_expired, ->() { with_artifacts_archive.where('artifacts_expire_at IS NULL OR artifacts_expire_at > ?', Time.now) }
     scope :with_expired_artifacts, ->() { with_artifacts_archive.where('artifacts_expire_at < ?', Time.now) }

changelogs/unreleased/add-background-migrations-for-not-archived-traces.yml

+---
+title: Add background migrations for archiving legacy job traces
+merge_request: 19194
+author:
+type: performance

db/post_migrate/20180529152628_schedule_to_archive_legacy_traces.rb

+class ScheduleToArchiveLegacyTraces < ActiveRecord::Migration
+  include Gitlab::Database::MigrationHelpers
+
+  DOWNTIME = false
+  BATCH_SIZE = 5000
+  BACKGROUND_MIGRATION_CLASS = 'ArchiveLegacyTraces'
+
+  disable_ddl_transaction!
+
+  class Build < ActiveRecord::Base
+    include EachBatch
+    self.table_name = 'ci_builds'
+    self.inheritance_column = :_type_disabled # Disable STI
+
+    scope :type_build, -> { where(type: 'Ci::Build') }
+
+    scope :finished, -> { where(status: [:success, :failed, :canceled]) }
+
+    scope :without_archived_trace, -> do
+      where('NOT EXISTS (SELECT 1 FROM ci_job_artifacts WHERE ci_builds.id = ci_job_artifacts.job_id AND ci_job_artifacts.file_type = 3)')
+    end
+  end
+
+  def up
+    queue_background_migration_jobs_by_range_at_intervals(
+      ::ScheduleToArchiveLegacyTraces::Build.type_build.finished.without_archived_trace,
+      BACKGROUND_MIGRATION_CLASS,
+      5.minutes,
+      batch_size: BATCH_SIZE)
+  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: 20180529093006) do
+ActiveRecord::Schema.define(version: 20180529152628) do
   # These are extensions that must be enabled in order to support this database
   enable_extension "plpgsql"

lib/gitlab/background_migration/archive_legacy_traces.rb

+# frozen_string_literal: true
+# rubocop:disable Metrics/AbcSize
+# rubocop:disable Style/Documentation
+
+module Gitlab
+  module BackgroundMigration
+    class ArchiveLegacyTraces
+      def perform(start_id, stop_id)
+        # This background migration directly refers to ::Ci::Build model which is defined in application code.
+        # In general, migration code should be isolated as much as possible in order to be idempotent.
+        # However, `archive!` method is too complicated to be replicated by coping its subsequent code.
+        # So we chose a way to use ::Ci::Build directly and we don't change the `archive!` method until 11.1
+        ::Ci::Build.finished.without_archived_trace
+          .where(id: start_id..stop_id).find_each do |build|
+            begin
+              build.trace.archive!
+            rescue => e
+              Rails.logger.error "Failed to archive live trace. id: #{build.id} message: #{e.message}"
+            end
+          end
+      end
+    end
+  end
+end

lib/tasks/gitlab/traces.rake

@@ -8,9 +8,7 @@ namespace :gitlab do
       logger = Logger.new(STDOUT)
       logger.info('Archiving legacy traces')
-      Ci::Build.finished
-        .where('NOT EXISTS (?)',
-          Ci::JobArtifact.select(1).trace.where('ci_builds.id = ci_job_artifacts.job_id'))
+      Ci::Build.finished.without_archived_trace
         .order(id: :asc)
         .find_in_batches(batch_size: 1000) do |jobs|
         job_ids = jobs.map { |job| [job.id] }

spec/lib/gitlab/background_migration/archive_legacy_traces_spec.rb

+require 'spec_helper'
+
+describe Gitlab::BackgroundMigration::ArchiveLegacyTraces, :migration, schema: 20180529152628 do
+  include TraceHelpers
+
+  let(:namespaces) { table(:namespaces) }
+  let(:projects) { table(:projects) }
+  let(:builds) { table(:ci_builds) }
+  let(:job_artifacts) { table(:ci_job_artifacts) }
+
+  before do
+    namespaces.create!(id: 123, name: 'gitlab1', path: 'gitlab1')
+    projects.create!(id: 123, name: 'gitlab1', path: 'gitlab1', namespace_id: 123)
+    @build = builds.create!(id: 1, project_id: 123, status: 'success', type: 'Ci::Build')
+  end
+
+  context 'when trace file exsits at the right place' do
+    before do
+      create_legacy_trace(@build, 'trace in file')
+    end
+
+    it 'correctly archive legacy traces' do
+      expect(job_artifacts.count).to eq(0)
+      expect(File.exist?(legacy_trace_path(@build))).to be_truthy
+
+      described_class.new.perform(1, 1)
+
+      expect(job_artifacts.count).to eq(1)
+      expect(File.exist?(legacy_trace_path(@build))).to be_falsy
+      expect(File.read(archived_trace_path(job_artifacts.first))).to eq('trace in file')
+    end
+  end
+
+  context 'when trace file does not exsits at the right place' do
+    it 'does not raise errors nor create job artifact' do
+      expect { described_class.new.perform(1, 1) }.not_to raise_error
+
+      expect(job_artifacts.count).to eq(0)
+    end
+  end
+
+  context 'when trace data exsits in database' do
+    before do
+      create_legacy_trace_in_db(@build, 'trace in db')
+    end
+
+    it 'correctly archive legacy traces' do
+      expect(job_artifacts.count).to eq(0)
+      expect(@build.read_attribute(:trace)).not_to be_empty
+
+      described_class.new.perform(1, 1)
+
+      @build.reload
+      expect(job_artifacts.count).to eq(1)
+      expect(@build.read_attribute(:trace)).to be_nil
+      expect(File.read(archived_trace_path(job_artifacts.first))).to eq('trace in db')
+    end
+  end
+end

spec/migrations/schedule_to_archive_legacy_traces_spec.rb

+require 'spec_helper'
+require Rails.root.join('db', 'post_migrate', '20180529152628_schedule_to_archive_legacy_traces')
+
+describe ScheduleToArchiveLegacyTraces, :migration do
+  include TraceHelpers
+
+  let(:namespaces) { table(:namespaces) }
+  let(:projects) { table(:projects) }
+  let(:builds) { table(:ci_builds) }
+  let(:job_artifacts) { table(:ci_job_artifacts) }
+
+  before do
+    namespaces.create!(id: 123, name: 'gitlab1', path: 'gitlab1')
+    projects.create!(id: 123, name: 'gitlab1', path: 'gitlab1', namespace_id: 123)
+    @build_success = builds.create!(id: 1, project_id: 123, status: 'success', type: 'Ci::Build')
+    @build_failed = builds.create!(id: 2, project_id: 123, status: 'failed', type: 'Ci::Build')
+    @builds_canceled = builds.create!(id: 3, project_id: 123, status: 'canceled', type: 'Ci::Build')
+    @build_running = builds.create!(id: 4, project_id: 123, status: 'running', type: 'Ci::Build')
+
+    create_legacy_trace(@build_success, 'This job is done')
+    create_legacy_trace(@build_failed, 'This job is done')
+    create_legacy_trace(@builds_canceled, 'This job is done')
+    create_legacy_trace(@build_running, 'This job is not done yet')
+  end
+
+  it 'correctly archive legacy traces' do
+    expect(job_artifacts.count).to eq(0)
+    expect(File.exist?(legacy_trace_path(@build_success))).to be_truthy
+    expect(File.exist?(legacy_trace_path(@build_failed))).to be_truthy
+    expect(File.exist?(legacy_trace_path(@builds_canceled))).to be_truthy
+    expect(File.exist?(legacy_trace_path(@build_running))).to be_truthy
+
+    migrate!
+
+    expect(job_artifacts.count).to eq(3)
+    expect(File.exist?(legacy_trace_path(@build_success))).to be_falsy
+    expect(File.exist?(legacy_trace_path(@build_failed))).to be_falsy
+    expect(File.exist?(legacy_trace_path(@builds_canceled))).to be_falsy
+    expect(File.exist?(legacy_trace_path(@build_running))).to be_truthy
+    expect(File.exist?(archived_trace_path(job_artifacts.where(job_id: @build_success.id).first))).to be_truthy
+    expect(File.exist?(archived_trace_path(job_artifacts.where(job_id: @build_failed.id).first))).to be_truthy
+    expect(File.exist?(archived_trace_path(job_artifacts.where(job_id: @builds_canceled.id).first))).to be_truthy
+    expect(job_artifacts.where(job_id: @build_running.id)).not_to be_exist
+  end
+end

spec/support/trace/trace_helpers.rb

+module TraceHelpers
+  def create_legacy_trace(build, content)
+    File.open(legacy_trace_path(build), 'wb') { |stream| stream.write(content) }
+  end
+
+  def create_legacy_trace_in_db(build, content)
+    build.update_column(:trace, content)
+  end
+
+  def legacy_trace_path(build)
+    legacy_trace_dir = File.join(Settings.gitlab_ci.builds_path,
+      build.created_at.utc.strftime("%Y_%m"),
+      build.project_id.to_s)
+
+    FileUtils.mkdir_p(legacy_trace_dir)
+
+    File.join(legacy_trace_dir, "#{build.id}.log")
+  end
+
+  def archived_trace_path(job_artifact)
+    disk_hash = Digest::SHA2.hexdigest(job_artifact.project_id.to_s)
+    creation_date = job_artifact.created_at.utc.strftime('%Y_%m_%d')
+
+    File.join(Gitlab.config.artifacts.path, disk_hash[0..1], disk_hash[2..3], disk_hash,
+      creation_date, job_artifact.job_id.to_s, job_artifact.id.to_s, 'job.log')
+  end
+end