activerecord/lib/active_record/connection_handling.rb
# frozen_string_literal: true
module ActiveRecord
# = Active Record Connection Handling
module ConnectionHandling
RAILS_ENV = -> { (Rails.env if defined?(Rails.env)) || ENV["RAILS_ENV"].presence || ENV["RACK_ENV"].presence }
DEFAULT_ENV = -> { RAILS_ENV.call || "default_env" }
# Establishes the connection to the database. Accepts a hash as input where
# the <tt>:adapter</tt> key must be specified with the name of a database adapter (in lower-case)
# example for regular databases (MySQL, PostgreSQL, etc):
#
# ActiveRecord::Base.establish_connection(
# adapter: "mysql2",
# host: "localhost",
# username: "myuser",
# password: "mypass",
# database: "somedatabase"
# )
#
# Example for SQLite database:
#
# ActiveRecord::Base.establish_connection(
# adapter: "sqlite3",
# database: "path/to/dbfile"
# )
#
# Also accepts keys as strings (for parsing from YAML for example):
#
# ActiveRecord::Base.establish_connection(
# "adapter" => "sqlite3",
# "database" => "path/to/dbfile"
# )
#
# Or a URL:
#
# ActiveRecord::Base.establish_connection(
# "postgres://myuser:mypass@localhost/somedatabase"
# )
#
# In case {ActiveRecord::Base.configurations}[rdoc-ref:Core.configurations]
# is set (\Rails automatically loads the contents of config/database.yml into it),
# a symbol can also be given as argument, representing a key in the
# configuration hash:
#
# ActiveRecord::Base.establish_connection(:production)
#
# The exceptions AdapterNotSpecified, AdapterNotFound, and +ArgumentError+
# may be returned on an error.
def establish_connection(config_or_env = nil)
config_or_env ||= DEFAULT_ENV.call.to_sym
db_config = resolve_config_for_connection(config_or_env)
connection_handler.establish_connection(db_config, owner_name: self, role: current_role, shard: current_shard)
end
# Connects a model to the databases specified. The +database+ keyword
# takes a hash consisting of a +role+ and a +database_key+.
#
# This will look up the database config using the +database_key+ and
# establish a connection to that config.
#
# class AnimalsModel < ApplicationRecord
# self.abstract_class = true
#
# connects_to database: { writing: :primary, reading: :primary_replica }
# end
#
# +connects_to+ also supports horizontal sharding. The horizontal sharding API
# supports read replicas as well. You can connect a model to a list of shards like this:
#
# class AnimalsModel < ApplicationRecord
# self.abstract_class = true
#
# connects_to shards: {
# default: { writing: :primary, reading: :primary_replica },
# shard_two: { writing: :primary_shard_two, reading: :primary_shard_replica_two }
# }
# end
#
# Returns an array of database connections.
def connects_to(database: {}, shards: {})
raise NotImplementedError, "`connects_to` can only be called on ActiveRecord::Base or abstract classes" unless self == Base || abstract_class?
if database.present? && shards.present?
raise ArgumentError, "`connects_to` can only accept a `database` or `shards` argument, but not both arguments."
end
connections = []
if shards.empty?
shards[:default] = database
end
self.default_shard = shards.keys.first
shards.each do |shard, database_keys|
database_keys.each do |role, database_key|
db_config = resolve_config_for_connection(database_key)
self.connection_class = true
connections << connection_handler.establish_connection(db_config, owner_name: self, role: role, shard: shard.to_sym)
end
end
connections
end
# Connects to a role (e.g. writing, reading, or a custom role) and/or
# shard for the duration of the block. At the end of the block the
# connection will be returned to the original role / shard.
#
# If only a role is passed, Active Record will look up the connection
# based on the requested role. If a non-established role is requested
# an +ActiveRecord::ConnectionNotEstablished+ error will be raised:
#
# ActiveRecord::Base.connected_to(role: :writing) do
# Dog.create! # creates dog using dog writing connection
# end
#
# ActiveRecord::Base.connected_to(role: :reading) do
# Dog.create! # throws exception because we're on a replica
# end
#
# When swapping to a shard, the role must be passed as well. If a non-existent
# shard is passed, an +ActiveRecord::ConnectionNotEstablished+ error will be
# raised.
#
# When a shard and role is passed, Active Record will first lookup the role,
# and then look up the connection by shard key.
#
# ActiveRecord::Base.connected_to(role: :reading, shard: :shard_one_replica) do
# Dog.first # finds first Dog record stored on the shard one replica
# end
def connected_to(role: nil, shard: nil, prevent_writes: false, &blk)
if self != Base && !abstract_class
raise NotImplementedError, "calling `connected_to` is only allowed on ActiveRecord::Base or abstract classes."
end
if !connection_class? && !primary_class?
raise NotImplementedError, "calling `connected_to` is only allowed on the abstract class that established the connection."
end
unless role || shard
raise ArgumentError, "must provide a `shard` and/or `role`."
end
with_role_and_shard(role, shard, prevent_writes, &blk)
end
# Connects a role and/or shard to the provided connection names. Optionally +prevent_writes+
# can be passed to block writes on a connection. +reading+ will automatically set
# +prevent_writes+ to true.
#
# +connected_to_many+ is an alternative to deeply nested +connected_to+ blocks.
#
# Usage:
#
# ActiveRecord::Base.connected_to_many(AnimalsRecord, MealsRecord, role: :reading) do
# Dog.first # Read from animals replica
# Dinner.first # Read from meals replica
# Person.first # Read from primary writer
# end
def connected_to_many(*classes, role:, shard: nil, prevent_writes: false)
classes = classes.flatten
if self != Base || classes.include?(Base)
raise NotImplementedError, "connected_to_many can only be called on ActiveRecord::Base."
end
prevent_writes = true if role == ActiveRecord.reading_role
append_to_connected_to_stack(role: role, shard: shard, prevent_writes: prevent_writes, klasses: classes)
yield
ensure
connected_to_stack.pop
end
# Use a specified connection.
#
# This method is useful for ensuring that a specific connection is
# being used. For example, when booting a console in readonly mode.
#
# It is not recommended to use this method in a request since it
# does not yield to a block like +connected_to+.
def connecting_to(role: default_role, shard: default_shard, prevent_writes: false)
prevent_writes = true if role == ActiveRecord.reading_role
append_to_connected_to_stack(role: role, shard: shard, prevent_writes: prevent_writes, klasses: [self])
end
# Prohibit swapping shards while inside of the passed block.
#
# In some cases you may want to be able to swap shards but not allow a
# nested call to connected_to or connected_to_many to swap again. This
# is useful in cases you're using sharding to provide per-request
# database isolation.
def prohibit_shard_swapping(enabled = true)
prev_value = ActiveSupport::IsolatedExecutionState[:active_record_prohibit_shard_swapping]
ActiveSupport::IsolatedExecutionState[:active_record_prohibit_shard_swapping] = enabled
yield
ensure
ActiveSupport::IsolatedExecutionState[:active_record_prohibit_shard_swapping] = prev_value
end
# Determine whether or not shard swapping is currently prohibited
def shard_swapping_prohibited?
ActiveSupport::IsolatedExecutionState[:active_record_prohibit_shard_swapping]
end
# Prevent writing to the database regardless of role.
#
# In some cases you may want to prevent writes to the database
# even if you are on a database that can write. +while_preventing_writes+
# will prevent writes to the database for the duration of the block.
#
# This method does not provide the same protection as a readonly
# user and is meant to be a safeguard against accidental writes.
#
# See +READ_QUERY+ for the queries that are blocked by this
# method.
def while_preventing_writes(enabled = true, &block)
connected_to(role: current_role, prevent_writes: enabled, &block)
end
# Returns true if role is the current connected role and/or
# current connected shard. If no shard is passed, the default will be
# used.
#
# ActiveRecord::Base.connected_to(role: :writing) do
# ActiveRecord::Base.connected_to?(role: :writing) #=> true
# ActiveRecord::Base.connected_to?(role: :reading) #=> false
# end
#
# ActiveRecord::Base.connected_to(role: :reading, shard: :shard_one) do
# ActiveRecord::Base.connected_to?(role: :reading, shard: :shard_one) #=> true
# ActiveRecord::Base.connected_to?(role: :reading, shard: :default) #=> false
# ActiveRecord::Base.connected_to?(role: :writing, shard: :shard_one) #=> true
# end
def connected_to?(role:, shard: ActiveRecord::Base.default_shard)
current_role == role.to_sym && current_shard == shard.to_sym
end
# Clears the query cache for all connections associated with the current thread.
def clear_query_caches_for_current_thread
connection_handler.each_connection_pool do |pool|
pool.clear_query_cache
end
end
# Returns the connection currently associated with the class. This can
# also be used to "borrow" the connection to do database work unrelated
# to any of the specific Active Records.
# The connection will remain leased for the entire duration of the request
# or job, or until +#release_connection+ is called.
def lease_connection
connection_pool.lease_connection
end
# Soft deprecated. Use +#with_connection+ or +#lease_connection+ instead.
def connection
pool = connection_pool
if pool.permanent_lease?
case ActiveRecord.permanent_connection_checkout
when :deprecated
ActiveRecord.deprecator.warn <<~MESSAGE
Called deprecated `ActiveRecord::Base.connection` method.
Either use `with_connection` or `lease_connection`.
MESSAGE
when :disallowed
raise ActiveRecordError, <<~MESSAGE
Called deprecated `ActiveRecord::Base.connection` method.
Either use `with_connection` or `lease_connection`.
MESSAGE
end
pool.lease_connection
else
pool.active_connection
end
end
# Return the currently leased connection into the pool
def release_connection
connection_pool.release_connection
end
# Checkouts a connection from the pool, yield it and then check it back in.
# If a connection was already leased via #lease_connection or a parent call to
# #with_connection, that same connection is yieled.
# If #lease_connection is called inside the block, the connection won't be checked
# back in.
# If #connection is called inside the block, the connection won't be checked back in
# unless the +prevent_permanent_checkout+ argument is set to +true+.
def with_connection(prevent_permanent_checkout: false, &block)
connection_pool.with_connection(prevent_permanent_checkout: prevent_permanent_checkout, &block)
end
attr_writer :connection_specification_name
# Returns the connection specification name from the current class or its parent.
def connection_specification_name
if @connection_specification_name.nil?
return self == Base ? Base.name : superclass.connection_specification_name
end
@connection_specification_name
end
def primary_class? # :nodoc:
self == Base || application_record_class?
end
# Returns the db_config object from the associated connection:
#
# ActiveRecord::Base.connection_db_config
# #<ActiveRecord::DatabaseConfigurations::HashConfig:0x00007fd1acbded10 @env_name="development",
# @name="primary", @config={pool: 5, timeout: 5000, database: "storage/development.sqlite3", adapter: "sqlite3"}>
#
# Use only for reading.
def connection_db_config
connection_pool.db_config
end
def adapter_class # :nodoc:
connection_pool.db_config.adapter_class
end
def connection_pool
connection_handler.retrieve_connection_pool(connection_specification_name, role: current_role, shard: current_shard, strict: true)
end
def retrieve_connection
connection_handler.retrieve_connection(connection_specification_name, role: current_role, shard: current_shard)
end
# Returns +true+ if Active Record is connected.
def connected?
connection_handler.connected?(connection_specification_name, role: current_role, shard: current_shard)
end
def remove_connection
name = @connection_specification_name if defined?(@connection_specification_name)
# if removing a connection that has a pool, we reset the
# connection_specification_name so it will use the parent
# pool.
if connection_handler.retrieve_connection_pool(name, role: current_role, shard: current_shard)
self.connection_specification_name = nil
end
connection_handler.remove_connection_pool(name, role: current_role, shard: current_shard)
end
def schema_cache # :nodoc:
connection_pool.schema_cache
end
def clear_cache! # :nodoc:
connection_pool.schema_cache.clear!
end
private
def resolve_config_for_connection(config_or_env)
raise "Anonymous class is not allowed." unless name
connection_name = primary_class? ? Base.name : name
self.connection_specification_name = connection_name
Base.configurations.resolve(config_or_env)
end
def with_role_and_shard(role, shard, prevent_writes)
prevent_writes = true if role == ActiveRecord.reading_role
append_to_connected_to_stack(role: role, shard: shard, prevent_writes: prevent_writes, klasses: [self])
return_value = yield
return_value.load if return_value.is_a? ActiveRecord::Relation
return_value
ensure
self.connected_to_stack.pop
end
def append_to_connected_to_stack(entry)
if shard_swapping_prohibited? && entry[:shard].present?
raise ArgumentError, "cannot swap `shard` while shard swapping is prohibited."
end
connected_to_stack << entry
end
end
end