Merge branch '340832-reindex-for-multiple-databases' into 'master'

Update gitlab:db:reindex to work with multiple databases

See merge request gitlab-org/gitlab!72745

.rubocop_manual_todo.yml

@@ -2449,8 +2449,6 @@ Database/MultipleDatabases:
     - 'lib/gitlab/database/postgresql_adapter/dump_schema_versions_mixin.rb'
     - 'lib/gitlab/database/postgresql_database_tasks/load_schema_versions_mixin.rb'
     - 'lib/gitlab/database.rb'
-    - 'lib/gitlab/database/reindexing/concurrent_reindex.rb'
-    - 'lib/gitlab/database/reindexing/reindex_concurrently.rb'
     - 'lib/gitlab/database/schema_cache_with_renamed_table.rb'
     - 'lib/gitlab/database/schema_migrations/context.rb'
     - 'lib/gitlab/database/schema_version_files.rb'

db/post_migrate/20210824174615_prepare_ci_builds_metadata_and_ci_build_async_indexes.rb

@@ -42,7 +42,7 @@ class PrepareCiBuildsMetadataAndCiBuildAsyncIndexes < ActiveRecord::Migration[6.
 
     return if index_name_exists?(table_name, index_name)
 
-    async_index = Gitlab::Database::AsyncIndexes::PostgresAsyncIndex.safe_find_or_create_by!(name: index_name) do |rec|
+    async_index = Gitlab::Database::AsyncIndexes::PostgresAsyncIndex.find_or_create_by!(name: index_name) do |rec|
       rec.table_name = table_name
       rec.definition = definition
     end

lib/gitlab/database/async_indexes/index_creator.rb

@@ -40,7 +40,7 @@ module Gitlab
         end
 
         def connection
-          @connection ||= ApplicationRecord.connection
+          @connection ||= async_index.connection
         end
 
         def lease_timeout

lib/gitlab/database/async_indexes/postgres_async_index.rb

@@ -3,7 +3,7 @@
 module Gitlab
   module Database
     module AsyncIndexes
-      class PostgresAsyncIndex < ApplicationRecord
+      class PostgresAsyncIndex < SharedModel
         self.table_name = 'postgres_async_indexes'
 
         MAX_IDENTIFIER_LENGTH = Gitlab::Database::MigrationHelpers::MAX_IDENTIFIER_NAME_LENGTH

lib/gitlab/database/postgres_index.rb

@@ -2,7 +2,7 @@
 
 module Gitlab
   module Database
-    class PostgresIndex < ActiveRecord::Base
+    class PostgresIndex < SharedModel
       include Gitlab::Utils::StrongMemoize
 
       self.table_name = 'postgres_indexes'

lib/gitlab/database/postgres_index_bloat_estimate.rb

@@ -6,7 +6,7 @@ module Gitlab
     # for all indexes can be expensive in a large database.
     #
     # Best used on a per-index basis.
-    class PostgresIndexBloatEstimate < ActiveRecord::Base
+    class PostgresIndexBloatEstimate < SharedModel
       self.table_name = 'postgres_index_bloat_estimates'
       self.primary_key = 'identifier'
 

lib/gitlab/database/reindexing.rb

@@ -27,13 +27,14 @@ module Gitlab
           Gitlab::AppLogger.info("Removing index #{index.identifier} which is a leftover, temporary index from previous reindexing activity")
 
           retries = Gitlab::Database::WithLockRetriesOutsideTransaction.new(
+            connection: index.connection,
             timing_configuration: REMOVE_INDEX_RETRY_CONFIG,
             klass: self.class,
             logger: Gitlab::AppLogger
           )
 
           retries.run(raise_on_exhaustion: false) do
-            ApplicationRecord.connection.tap do |conn|
+            index.connection.tap do |conn|
               conn.execute("DROP INDEX CONCURRENTLY IF EXISTS #{conn.quote_table_name(index.schema)}.#{conn.quote_table_name(index.name)}")
             end
           end

lib/gitlab/database/reindexing/reindex_action.rb

@@ -3,7 +3,7 @@
 module Gitlab
   module Database
     module Reindexing
-      class ReindexAction < ActiveRecord::Base
+      class ReindexAction < SharedModel
         self.table_name = 'postgres_reindex_actions'
 
         belongs_to :index, foreign_key: :index_identifier, class_name: 'Gitlab::Database::PostgresIndex'

lib/gitlab/database/reindexing/reindex_concurrently.rb

@@ -99,6 +99,7 @@ module Gitlab
           logger.info("Removing dangling index #{index.identifier}")
 
           retries = Gitlab::Database::WithLockRetriesOutsideTransaction.new(
+            connection: connection,
             timing_configuration: REMOVE_INDEX_RETRY_CONFIG,
             klass: self.class,
             logger: logger
@@ -109,11 +110,6 @@ module Gitlab
           end
         end
 
-        def with_lock_retries(&block)
-          arguments = { klass: self.class, logger: logger }
-          Gitlab::Database::WithLockRetries.new(**arguments).run(raise_on_exhaustion: true, &block)
-        end
-
         def set_statement_timeout
           execute("SET statement_timeout TO '%ds'" % STATEMENT_TIMEOUT)
           yield
@@ -123,7 +119,7 @@ module Gitlab
 
         delegate :execute, :quote_table_name, to: :connection
         def connection
-          @connection ||= ActiveRecord::Base.connection
+          @connection ||= index.connection
         end
       end
     end

lib/tasks/gitlab/db.rake

@@ -161,31 +161,33 @@ namespace :gitlab do
     end
 
     desc 'reindex a regular index without downtime to eliminate bloat'
-    task :reindex, [:index_name] => :environment do |_, args|
+    task :reindex, [:index_name, :database] => :environment do |_, args|
       unless Feature.enabled?(:database_reindexing, type: :ops)
         puts "This feature (database_reindexing) is currently disabled.".color(:yellow)
         exit
       end
 
-      indexes = Gitlab::Database::PostgresIndex.reindexing_support
+      Gitlab::Database::EachDatabase.each_database_connection do |connection, connection_name|
+        indexes = Gitlab::Database::PostgresIndex.reindexing_support
 
-      if identifier = args[:index_name]
-        raise ArgumentError, "Index name is not fully qualified with a schema: #{identifier}" unless identifier =~ /^\w+\.\w+$/
+        if (identifier = args[:index_name]) && (args.fetch(:database, 'main') == connection_name)
+          raise ArgumentError, "Index name is not fully qualified with a schema: #{identifier}" unless identifier =~ /^\w+\.\w+$/
 
-        indexes = indexes.where(identifier: identifier)
+          indexes = indexes.where(identifier: identifier)
 
-        raise "Index not found or not supported: #{args[:index_name]}" if indexes.empty?
-      end
+          raise "Index #{args[:index_name]} for #{connection_name} database not found or not supported" if indexes.empty?
+        end
 
-      ActiveRecord::Base.logger = Logger.new($stdout) if Gitlab::Utils.to_boolean(ENV['LOG_QUERIES_TO_CONSOLE'], default: false)
+        Gitlab::Database::SharedModel.logger = Logger.new($stdout) if Gitlab::Utils.to_boolean(ENV['LOG_QUERIES_TO_CONSOLE'], default: false)
 
-      # Cleanup leftover temporary indexes from previous, possibly aborted runs (if any)
-      Gitlab::Database::Reindexing.cleanup_leftovers!
+        # Cleanup leftover temporary indexes from previous, possibly aborted runs (if any)
+        Gitlab::Database::Reindexing.cleanup_leftovers!
 
-      # Hack: Before we do actual reindexing work, create async indexes
-      Gitlab::Database::AsyncIndexes.create_pending_indexes! if Feature.enabled?(:database_async_index_creation, type: :ops)
+        # Hack: Before we do actual reindexing work, create async indexes
+        Gitlab::Database::AsyncIndexes.create_pending_indexes! if Feature.enabled?(:database_async_index_creation, type: :ops)
 
-      Gitlab::Database::Reindexing.perform(indexes)
+        Gitlab::Database::Reindexing.perform(indexes)
+      end
     rescue StandardError => e
       Gitlab::AppLogger.error(e)
       raise

spec/lib/gitlab/database/async_indexes/postgres_async_index_spec.rb

@@ -3,6 +3,8 @@
 require 'spec_helper'
 
 RSpec.describe Gitlab::Database::AsyncIndexes::PostgresAsyncIndex, type: :model do
+  it { is_expected.to be_a Gitlab::Database::SharedModel }
+
   describe 'validations' do
     let(:identifier_limit) { described_class::MAX_IDENTIFIER_LENGTH }
     let(:definition_limit) { described_class::MAX_DEFINITION_LENGTH }

spec/lib/gitlab/database/postgres_index_bloat_estimate_spec.rb

@@ -13,6 +13,8 @@ RSpec.describe Gitlab::Database::PostgresIndexBloatEstimate do
 
   let(:identifier) { 'public.schema_migrations_pkey' }
 
+  it { is_expected.to be_a Gitlab::Database::SharedModel }
+
   describe '#bloat_size' do
     it 'returns the bloat size in bytes' do
       # We cannot reach much more about the bloat size estimate here

spec/lib/gitlab/database/postgres_index_spec.rb

@@ -22,6 +22,8 @@ RSpec.describe Gitlab::Database::PostgresIndex do
 
   it_behaves_like 'a postgres model'
 
+  it { is_expected.to be_a Gitlab::Database::SharedModel }
+
   describe '.reindexing_support' do
     it 'only non partitioned indexes' do
       expect(described_class.reindexing_support).to all(have_attributes(partitioned: false))

spec/lib/gitlab/database/reindexing/reindex_action_spec.rb

@@ -11,6 +11,8 @@ RSpec.describe Gitlab::Database::Reindexing::ReindexAction do
     swapout_view_for_table(:postgres_indexes)
   end
 
+  it { is_expected.to be_a Gitlab::Database::SharedModel }
+
   describe '.create_for' do
     subject { described_class.create_for(index) }
 

spec/tasks/gitlab/db_rake_spec.rb

@@ -252,10 +252,26 @@ RSpec.describe 'gitlab:db namespace rake task', :silence_stdout do
         run_rake_task('gitlab:db:reindex', '[public.foo_idx]')
       end
 
+      context 'when database name is provided' do
+        it 'calls the index rebuilder with the proper arguments when the database name match' do
+          allow(indexes).to receive(:where).with(identifier: 'public.foo_idx').and_return([index])
+          expect(Gitlab::Database::Reindexing).to receive(:perform).with([index])
+
+          run_rake_task('gitlab:db:reindex', '[public.foo_idx,main]')
+        end
+
+        it 'ignores the index and uses all candidate indexes if database name does not match' do
+          expect(Gitlab::Database::PostgresIndex).to receive(:reindexing_support).and_return(indexes)
+          expect(Gitlab::Database::Reindexing).to receive(:perform).with(indexes)
+
+          run_rake_task('gitlab:db:reindex', '[public.foo_idx,ci]')
+        end
+      end
+
       it 'raises an error if the index does not exist' do
         allow(indexes).to receive(:where).with(identifier: 'public.absent_index').and_return([])
 
-        expect { run_rake_task('gitlab:db:reindex', '[public.absent_index]') }.to raise_error(/Index not found/)
+        expect { run_rake_task('gitlab:db:reindex', '[public.absent_index]') }.to raise_error(/Index public.absent_index for main database not found or not supported/)
       end
 
       it 'raises an error if the index is not fully qualified with a schema' do