debian-mirror-gitlab/lib/gitlab/database.rb

378 lines
15 KiB
Ruby
Raw Normal View History

2018-12-13 13:39:08 +05:30
# frozen_string_literal: true
2015-10-24 18:46:33 +05:30
module Gitlab
module Database
2021-11-11 11:23:49 +05:30
DATABASE_NAMES = %w[main ci].freeze
MAIN_DATABASE_NAME = 'main'
2021-09-30 23:02:18 +05:30
CI_DATABASE_NAME = 'ci'
2021-11-11 11:23:49 +05:30
DEFAULT_POOL_HEADROOM = 10
2021-09-30 23:02:18 +05:30
2021-06-08 01:23:25 +05:30
# This constant is used when renaming tables concurrently.
# If you plan to rename a table using the `rename_table_safely` method, add your table here one milestone before the rename.
# Example:
# TABLES_TO_BE_RENAMED = {
# 'old_name' => 'new_name'
# }.freeze
2021-09-30 23:02:18 +05:30
TABLES_TO_BE_RENAMED = {}.freeze
2021-06-08 01:23:25 +05:30
2020-07-28 23:09:34 +05:30
# Minimum PostgreSQL version requirement per documentation:
# https://docs.gitlab.com/ee/install/requirements.html#postgresql-requirements
2021-09-04 01:27:46 +05:30
MINIMUM_POSTGRES_VERSION = 12
2020-07-28 23:09:34 +05:30
2016-06-22 15:30:34 +05:30
# https://www.postgresql.org/docs/9.2/static/datatype-numeric.html
MAX_INT_VALUE = 2147483647
2020-10-24 23:57:45 +05:30
MIN_INT_VALUE = -2147483648
2019-10-12 21:52:04 +05:30
2018-03-17 18:26:18 +05:30
# The max value between MySQL's TIMESTAMP and PostgreSQL's timestampz:
# https://www.postgresql.org/docs/9.1/static/datatype-datetime.html
# https://dev.mysql.com/doc/refman/5.7/en/datetime.html
2019-10-12 21:52:04 +05:30
# FIXME: this should just be the max value of timestampz
2018-03-17 18:26:18 +05:30
MAX_TIMESTAMP_VALUE = Time.at((1 << 31) - 1).freeze
2016-06-22 15:30:34 +05:30
2019-09-04 21:01:54 +05:30
# The maximum number of characters for text fields, to avoid DoS attacks via parsing huge text fields
2019-12-04 20:38:33 +05:30
# https://gitlab.com/gitlab-org/gitlab-foss/issues/61974
2019-09-04 21:01:54 +05:30
MAX_TEXT_SIZE_LIMIT = 1_000_000
2019-09-30 21:07:59 +05:30
# Minimum schema version from which migrations are supported
# Migrations before this version may have been removed
MIN_SCHEMA_VERSION = 20190506135400
MIN_SCHEMA_GITLAB_VERSION = '11.11.0'
2020-07-28 23:09:34 +05:30
# Schema we store dynamically managed partitions in (e.g. for time partitioning)
DYNAMIC_PARTITIONS_SCHEMA = :gitlab_partitions_dynamic
# Schema we store static partitions in (e.g. for hash partitioning)
STATIC_PARTITIONS_SCHEMA = :gitlab_partitions_static
# This is an extensive list of postgres schemas owned by GitLab
# It does not include the default public schema
EXTRA_SCHEMAS = [DYNAMIC_PARTITIONS_SCHEMA, STATIC_PARTITIONS_SCHEMA].freeze
2022-07-16 23:28:13 +05:30
PRIMARY_DATABASE_NAME = ActiveRecord::Base.connection_db_config.name.to_sym # rubocop:disable Database/MultipleDatabases
2021-06-08 01:23:25 +05:30
2021-11-18 22:05:49 +05:30
def self.database_base_models
@database_base_models ||= {
# Note that we use ActiveRecord::Base here and not ApplicationRecord.
# This is deliberate, as we also use these classes to apply load
# balancing to, and the load balancer must be enabled for _all_ models
# that inher from ActiveRecord::Base; not just our own models that
# inherit from ApplicationRecord.
main: ::ActiveRecord::Base,
2021-12-11 22:18:48 +05:30
ci: ::Ci::ApplicationRecord.connection_class? ? ::Ci::ApplicationRecord : nil
}.compact.with_indifferent_access.freeze
2017-08-17 22:00:37 +05:30
end
2022-08-13 15:12:31 +05:30
# This returns a list of databases that contains all the gitlab_shared schema
# tables. We can't reuse database_base_models because Geo does not support
# the gitlab_shared tables yet.
def self.database_base_models_with_gitlab_shared
@database_base_models_with_gitlab_shared ||= {
# Note that we use ActiveRecord::Base here and not ApplicationRecord.
# This is deliberate, as we also use these classes to apply load
# balancing to, and the load balancer must be enabled for _all_ models
# that inher from ActiveRecord::Base; not just our own models that
# inherit from ApplicationRecord.
main: ::ActiveRecord::Base,
ci: ::Ci::ApplicationRecord.connection_class? ? ::Ci::ApplicationRecord : nil
}.compact.with_indifferent_access.freeze
end
# This returns a list of databases whose connection supports database load
# balancing. We can't reuse the database_base_models method because the Geo
# database does not support load balancing yet.
#
# TODO: https://gitlab.com/gitlab-org/geo-team/discussions/-/issues/5032
def self.database_base_models_using_load_balancing
@database_base_models_with_gitlab_shared ||= {
# Note that we use ActiveRecord::Base here and not ApplicationRecord.
# This is deliberate, as we also use these classes to apply load
# balancing to, and the load balancer must be enabled for _all_ models
# that inher from ActiveRecord::Base; not just our own models that
# inherit from ApplicationRecord.
main: ::ActiveRecord::Base,
ci: ::Ci::ApplicationRecord.connection_class? ? ::Ci::ApplicationRecord : nil
}.compact.with_indifferent_access.freeze
end
2022-01-26 12:08:38 +05:30
# This returns a list of base models with connection associated for a given gitlab_schema
def self.schemas_to_base_models
@schemas_to_base_models ||= {
gitlab_main: [self.database_base_models.fetch(:main)],
gitlab_ci: [self.database_base_models[:ci] || self.database_base_models.fetch(:main)], # use CI or fallback to main
2022-08-13 15:12:31 +05:30
gitlab_shared: database_base_models_with_gitlab_shared.values, # all models
gitlab_internal: database_base_models.values # all models
2022-01-26 12:08:38 +05:30
}.with_indifferent_access.freeze
end
def self.all_database_names
DATABASE_NAMES
end
2021-11-11 11:23:49 +05:30
# We configure the database connection pool size automatically based on the
# configured concurrency. We also add some headroom, to make sure we don't
# run out of connections when more threads besides the 'user-facing' ones
# are running.
#
# Read more about this in
# doc/development/database/client_side_connection_pool.md
def self.default_pool_size
headroom =
(ENV["DB_POOL_HEADROOM"].presence || DEFAULT_POOL_HEADROOM).to_i
Gitlab::Runtime.max_threads + headroom
end
2021-09-30 23:02:18 +05:30
def self.has_config?(database_name)
Gitlab::Application.config.database_configuration[Rails.env].include?(database_name.to_s)
end
2022-04-04 11:22:00 +05:30
class PgUser < ApplicationRecord
self.table_name = 'pg_user'
self.primary_key = :usename
end
# rubocop: disable CodeReuse/ActiveRecord
def self.check_for_non_superuser
user = PgUser.find_by('usename = CURRENT_USER')
am_i_superuser = user.usesuper
Gitlab::AppLogger.info(
"Account details: User: \"#{user.usename}\", UseSuper: (#{am_i_superuser})"
)
raise 'Error: detected superuser' if am_i_superuser
rescue ActiveRecord::StatementInvalid
raise 'User CURRENT_USER not found'
end
# rubocop: enable CodeReuse/ActiveRecord
2020-07-28 23:09:34 +05:30
def self.check_postgres_version_and_print_warning
return if Gitlab::Runtime.rails_runner?
2021-12-11 22:18:48 +05:30
database_base_models.each do |name, model|
database = Gitlab::Database::Reflection.new(model)
next if database.postgresql_minimum_supported_version?
2021-10-27 15:23:28 +05:30
Kernel.warn ERB.new(Rainbow.new.wrap(<<~EOS).red).result
******************************************************************************
2021-12-11 22:18:48 +05:30
You are using PostgreSQL #{database.version} for the #{name} database, but PostgreSQL >= <%= Gitlab::Database::MINIMUM_POSTGRES_VERSION %>
2021-10-27 15:23:28 +05:30
is required for this version of GitLab.
<% if Rails.env.development? || Rails.env.test? %>
If using gitlab-development-kit, please find the relevant steps here:
https://gitlab.com/gitlab-org/gitlab-development-kit/-/blob/main/doc/howto/postgresql.md#upgrade-postgresql
<% end %>
Please upgrade your environment to a supported PostgreSQL version, see
https://docs.gitlab.com/ee/install/requirements.html#database for details.
******************************************************************************
EOS
rescue ActiveRecord::ActiveRecordError, PG::Error
# ignore - happens when Rake tasks yet have to create a database, e.g. for testing
end
2020-01-01 13:55:28 +05:30
end
def self.random
2019-10-12 21:52:04 +05:30
"RANDOM()"
end
2017-08-17 22:00:37 +05:30
def self.true_value
2019-10-12 21:52:04 +05:30
"'t'"
2015-12-23 02:04:40 +05:30
end
2017-08-17 22:00:37 +05:30
def self.false_value
2019-10-12 21:52:04 +05:30
"'f'"
2015-12-23 02:04:40 +05:30
end
2016-04-02 18:10:28 +05:30
2018-03-17 18:26:18 +05:30
def self.sanitize_timestamp(timestamp)
MAX_TIMESTAMP_VALUE > timestamp ? timestamp : MAX_TIMESTAMP_VALUE.dup
2017-09-10 17:25:29 +05:30
end
2022-05-07 20:08:51 +05:30
def self.all_uncached(&block)
# Calls to #uncached only disable caching for the current connection. Since the load balancer
# can potentially upgrade from read to read-write mode (using a different connection), we specify
# up-front that we'll explicitly use the primary for the duration of the operation.
Gitlab::Database::LoadBalancing::Session.current.use_primary do
2022-08-13 15:12:31 +05:30
base_models = database_base_models_using_load_balancing.values
2022-05-07 20:08:51 +05:30
base_models.reduce(block) { |blk, model| -> { model.uncached(&blk) } }.call
end
end
2021-10-27 15:23:28 +05:30
def self.allow_cross_joins_across_databases(url:)
# this method is implemented in:
# spec/support/database/prevent_cross_joins.rb
2021-11-11 11:23:49 +05:30
yield
2016-04-02 18:10:28 +05:30
end
2018-12-05 23:21:45 +05:30
def self.add_post_migrate_path_to_rails(force: false)
return if ENV['SKIP_POST_DEPLOYMENT_MIGRATIONS'] && !force
Rails.application.config.paths['db'].each do |db_path|
path = Rails.root.join(db_path, 'post_migrate').to_s
unless Rails.application.config.paths['db/migrate'].include? path
Rails.application.config.paths['db/migrate'] << path
# Rails memoizes migrations at certain points where it won't read the above
# path just yet. As such we must also update the following list of paths.
ActiveRecord::Migrator.migrations_paths << path
end
end
end
2019-09-30 21:07:59 +05:30
2021-10-27 15:23:28 +05:30
def self.db_config_names
2022-06-21 17:19:12 +05:30
::ActiveRecord::Base.configurations.configs_for(env_name: Rails.env).map(&:name) - ['geo']
2021-10-27 15:23:28 +05:30
end
2022-05-07 20:08:51 +05:30
# This returns all matching schemas that a given connection can use
# Since the `ActiveRecord::Base` might change the connection (from main to ci)
# This does not look at literal connection names, but rather compares
# models that are holders for a given db_config_name
def self.gitlab_schemas_for_connection(connection)
2022-07-23 23:45:48 +05:30
db_config = self.db_config_for_connection(connection)
# connection might not be yet adopted (returning NullPool, and no connection_klass)
# in such cases it is fine to ignore such connections
return unless db_config
primary_model = self.database_base_models.fetch(db_config.name.to_sym)
2022-06-21 17:19:12 +05:30
self.schemas_to_base_models.select do |_, child_models|
child_models.any? do |child_model|
child_model == primary_model || \
# The model might indicate a child connection, ensure that this is enclosed in a `db_config`
self.database_base_models[self.db_config_share_with(child_model.connection_db_config)] == primary_model
end
end.keys.map!(&:to_sym)
2022-05-07 20:08:51 +05:30
end
2021-11-11 11:23:49 +05:30
def self.db_config_for_connection(connection)
return unless connection
2021-09-30 23:02:18 +05:30
2022-07-23 23:45:48 +05:30
# For a ConnectionProxy we want to avoid ambiguous db_config as it may
# sometimes default to replica so we always return the primary config
# instead.
2022-05-07 20:08:51 +05:30
if connection.is_a?(::Gitlab::Database::LoadBalancing::ConnectionProxy)
2022-07-23 23:45:48 +05:30
return connection.load_balancer.configuration.db_config
2022-05-07 20:08:51 +05:30
end
2021-11-11 11:23:49 +05:30
# During application init we might receive `NullPool`
return unless connection.respond_to?(:pool) &&
connection.pool.respond_to?(:db_config)
connection.pool.db_config
end
# At the moment, the connection can only be retrieved by
# Gitlab::Database::LoadBalancer#read or #read_write or from the
# ActiveRecord directly. Therefore, if the load balancer doesn't
# recognize the connection, this method returns the primary role
# directly. In future, we may need to check for other sources.
# Expected returned names:
# main, main_replica, ci, ci_replica, unknown
def self.db_config_name(connection)
db_config = db_config_for_connection(connection)
db_config&.name || 'unknown'
2021-09-30 23:02:18 +05:30
end
2022-06-21 17:19:12 +05:30
# Currently the database configuration can only be shared with `main:`
# If the `database_tasks: false` is being used
# This is to be refined: https://gitlab.com/gitlab-org/gitlab/-/issues/356580
def self.db_config_share_with(db_config)
if db_config.database_tasks?
nil # no sharing
else
'main' # share with `main:`
end
end
2021-10-27 15:23:28 +05:30
def self.read_only?
false
2019-09-30 21:07:59 +05:30
end
2021-10-27 15:23:28 +05:30
def self.read_write?
!read_only?
2019-09-30 21:07:59 +05:30
end
# Monkeypatch rails with upgraded database observability
2021-10-27 15:23:28 +05:30
def self.install_transaction_metrics_patches!
2019-09-30 21:07:59 +05:30
ActiveRecord::Base.prepend(ActiveRecordBaseTransactionMetrics)
end
2021-10-27 15:23:28 +05:30
def self.install_transaction_context_patches!
ActiveRecord::ConnectionAdapters::TransactionManager
.prepend(TransactionManagerContext)
ActiveRecord::ConnectionAdapters::RealTransaction
.prepend(RealTransactionContext)
end
2019-09-30 21:07:59 +05:30
# MonkeyPatch for ActiveRecord::Base for adding observability
module ActiveRecordBaseTransactionMetrics
2021-04-17 20:07:23 +05:30
extend ActiveSupport::Concern
class_methods do
2022-08-13 15:12:31 +05:30
# A patch over ApplicationRecord.transaction that provides
2021-11-11 11:23:49 +05:30
# observability into transactional methods.
2021-04-17 20:07:23 +05:30
def transaction(**options, &block)
2021-12-11 22:18:48 +05:30
transaction_type = get_transaction_type(connection.transaction_open?, options[:requires_new])
::Gitlab::Database::Metrics.subtransactions_increment(self.name) if transaction_type == :sub_transaction
payload = { connection: connection, transaction_type: transaction_type }
2021-11-11 11:23:49 +05:30
2021-12-11 22:18:48 +05:30
ActiveSupport::Notifications.instrument('transaction.active_record', payload) do
2021-04-17 20:07:23 +05:30
super(**options, &block)
end
end
2021-12-11 22:18:48 +05:30
private
def get_transaction_type(transaction_open, requires_new_flag)
if transaction_open
return :sub_transaction if requires_new_flag
return :fake_transaction
end
:real_transaction
end
2019-09-30 21:07:59 +05:30
end
end
2021-10-27 15:23:28 +05:30
# rubocop:disable Gitlab/ModuleWithInstanceVariables
module TransactionManagerContext
def transaction_context
@stack.first.try(:gitlab_transaction_context)
end
end
module RealTransactionContext
def gitlab_transaction_context
@gitlab_transaction_context ||= ::Gitlab::Database::Transaction::Context.new
end
def commit
gitlab_transaction_context.commit
super
end
def rollback
gitlab_transaction_context.rollback
super
end
end
# rubocop:enable Gitlab/ModuleWithInstanceVariables
2015-10-24 18:46:33 +05:30
end
end
2019-12-04 20:38:33 +05:30
2021-06-08 01:23:25 +05:30
Gitlab::Database.prepend_mod_with('Gitlab::Database')