Active Record Migrations

Migrations can manage the evolution of a schema used by several physical databases. It’s a solution to the common problem of adding a field to make a new feature work in your local database, but being unsure of how to push that change to other developers and to the production server. With migrations, you can describe the transformations in self-contained classes that can be checked into version control systems and executed against another database that might be one, two, or five versions behind.

Example of a simple migration:

class AddSsl < ActiveRecord::Migration[7.1]
  def up
    add_column :accounts, :ssl_enabled, :boolean, default: true
  end

  def down
    remove_column :accounts, :ssl_enabled
  end
end

This migration will add a boolean flag to the accounts table and remove it if you’re backing out of the migration. It shows how all migrations have two methods up and down that describes the transformations required to implement or remove the migration. These methods can consist of both the migration specific methods like add_column and remove_column, but may also contain regular Ruby code for generating data needed for the transformations.

Example of a more complex migration that also needs to initialize data:

class AddSystemSettings < ActiveRecord::Migration[7.1]
  def up
    create_table :system_settings do |t|
      t.string  :name
      t.string  :label
      t.text    :value
      t.string  :type
      t.integer :position
    end

    SystemSetting.create  name:  'notice',
                          label: 'Use notice?',
                          value: 1
  end

  def down
    drop_table :system_settings
  end
end

This migration first adds the system_settings table, then creates the very first row in it using the Active Record model that relies on the table. It also uses the more advanced create_table syntax where you can specify a complete table schema in one block call.

Available transformations

Creation

  • create_join_table(table_1, table_2, options): Creates a join table having its name as the lexical order of the first two arguments. See ActiveRecord::ConnectionAdapters::SchemaStatements#create_join_table for details.

  • create_table(name, options): Creates a table called name and makes the table object available to a block that can then add columns to it, following the same format as add_column. See example above. The options hash is for fragments like “DEFAULT CHARSET=UTF-8” that are appended to the create table definition.

  • add_column(table_name, column_name, type, options): Adds a new column to the table called table_name named column_name specified to be one of the following types: :string, :text, :integer, :float, :decimal, :datetime, :timestamp, :time, :date, :binary, :boolean. A default value can be specified by passing an options hash like { default: 11 }. Other options include :limit and :null (e.g. { limit: 50, null: false }) – see ActiveRecord::ConnectionAdapters::TableDefinition#column for details.

  • add_foreign_key(from_table, to_table, options): Adds a new foreign key. from_table is the table with the key column, to_table contains the referenced primary key.

  • add_index(table_name, column_names, options): Adds a new index with the name of the column. Other options include :name, :unique (e.g. { name: 'users_name_index', unique: true }) and :order (e.g. { order: { name: :desc } }).

  • add_reference(:table_name, :reference_name): Adds a new column reference_name_id by default an integer. See ActiveRecord::ConnectionAdapters::SchemaStatements#add_reference for details.

  • add_timestamps(table_name, options): Adds timestamps (created_at and updated_at) columns to table_name.

Modification

  • change_column(table_name, column_name, type, options): Changes the column to a different type using the same parameters as add_column.

  • change_column_default(table_name, column_name, default_or_changes): Sets a default value for column_name defined by default_or_changes on table_name. Passing a hash containing :from and :to as default_or_changes will make this change reversible in the migration.

  • change_column_null(table_name, column_name, null, default = nil): Sets or removes a NOT NULL constraint on column_name. The null flag indicates whether the value can be NULL. See ActiveRecord::ConnectionAdapters::SchemaStatements#change_column_null for details.

  • change_table(name, options): Allows to make column alterations to the table called name. It makes the table object available to a block that can then add/remove columns, indexes, or foreign keys to it.

  • rename_column(table_name, column_name, new_column_name): Renames a column but keeps the type and content.

  • rename_index(table_name, old_name, new_name): Renames an index.

  • rename_table(old_name, new_name): Renames the table called old_name to new_name.

Deletion

  • drop_table(name): Drops the table called name.

  • drop_join_table(table_1, table_2, options): Drops the join table specified by the given arguments.

  • remove_column(table_name, column_name, type, options): Removes the column named column_name from the table called table_name.

  • remove_columns(table_name, *column_names): Removes the given columns from the table definition.

  • remove_foreign_key(from_table, to_table = nil, **options): Removes the given foreign key from the table called table_name.

  • remove_index(table_name, column: column_names): Removes the index specified by column_names.

  • remove_index(table_name, name: index_name): Removes the index specified by index_name.

  • remove_reference(table_name, ref_name, options): Removes the reference(s) on table_name specified by ref_name.

  • remove_timestamps(table_name, options): Removes the timestamp columns (created_at and updated_at) from the table definition.

Irreversible transformations

Some transformations are destructive in a manner that cannot be reversed. Migrations of that kind should raise an ActiveRecord::IrreversibleMigration exception in their down method.

Running migrations from within Rails

The Rails package has several tools to help create and apply migrations.

To generate a new migration, you can use

bin/rails generate migration MyNewMigration

where MyNewMigration is the name of your migration. The generator will create an empty migration file timestamp_my_new_migration.rb in the db/migrate/ directory where timestamp is the UTC formatted date and time that the migration was generated.

There is a special syntactic shortcut to generate migrations that add fields to a table.

bin/rails generate migration add_fieldname_to_tablename fieldname:string

This will generate the file timestamp_add_fieldname_to_tablename.rb, which will look like this:

class AddFieldnameToTablename < ActiveRecord::Migration[7.1]
  def change
    add_column :tablenames, :fieldname, :string
  end
end

To run migrations against the currently configured database, use bin/rails db:migrate. This will update the database by running all of the pending migrations, creating the schema_migrations table (see “About the schema_migrations table” section below) if missing. It will also invoke the db:schema:dump command, which will update your db/schema.rb file to match the structure of your database.

To roll the database back to a previous migration version, use bin/rails db:rollback VERSION=X where X is the version to which you wish to downgrade. Alternatively, you can also use the STEP option if you wish to rollback last few migrations. bin/rails db:rollback STEP=2 will rollback the latest two migrations.

If any of the migrations throw an ActiveRecord::IrreversibleMigration exception, that step will fail and you’ll have some manual work to do.

More examples

Not all migrations change the schema. Some just fix the data:

class RemoveEmptyTags < ActiveRecord::Migration[7.1]
  def up
    Tag.all.each { |tag| tag.destroy if tag.pages.empty? }
  end

  def down
    # not much we can do to restore deleted data
    raise ActiveRecord::IrreversibleMigration, "Can't recover the deleted tags"
  end
end

Others remove columns when they migrate up instead of down:

class RemoveUnnecessaryItemAttributes < ActiveRecord::Migration[7.1]
  def up
    remove_column :items, :incomplete_items_count
    remove_column :items, :completed_items_count
  end

  def down
    add_column :items, :incomplete_items_count
    add_column :items, :completed_items_count
  end
end

And sometimes you need to do something in SQL not abstracted directly by migrations:

class MakeJoinUnique < ActiveRecord::Migration[7.1]
  def up
    execute "ALTER TABLE `pages_linked_pages` ADD UNIQUE `page_id_linked_page_id` (`page_id`,`linked_page_id`)"
  end

  def down
    execute "ALTER TABLE `pages_linked_pages` DROP INDEX `page_id_linked_page_id`"
  end
end

Using a model after changing its table

Sometimes you’ll want to add a column in a migration and populate it immediately after. In that case, you’ll need to make a call to Base#reset_column_information in order to ensure that the model has the latest column data from after the new column was added. Example:

class AddPeopleSalary < ActiveRecord::Migration[7.1]
  def up
    add_column :people, :salary, :integer
    Person.reset_column_information
    Person.all.each do |p|
      p.update_attribute :salary, SalaryCalculator.compute(p)
    end
  end
end

Controlling verbosity

By default, migrations will describe the actions they are taking, writing them to the console as they happen, along with benchmarks describing how long each step took.

You can quiet them down by setting ActiveRecord::Migration.verbose = false.

You can also insert your own messages and benchmarks by using the say_with_time method:

def up
  ...
  say_with_time "Updating salaries..." do
    Person.all.each do |p|
      p.update_attribute :salary, SalaryCalculator.compute(p)
    end
  end
  ...
end

The phrase “Updating salaries…” would then be printed, along with the benchmark for the block when the block completes.

Timestamped Migrations

By default, Rails generates migrations that look like:

20080717013526_your_migration_name.rb

The prefix is a generation timestamp (in UTC).

If you’d prefer to use numeric prefixes, you can turn timestamped migrations off by setting:

config.active_record.timestamped_migrations = false

In application.rb.

Reversible Migrations

Reversible migrations are migrations that know how to go down for you. You simply supply the up logic, and the Migration system figures out how to execute the down commands for you.

To define a reversible migration, define the change method in your migration like this:

class TenderloveMigration < ActiveRecord::Migration[7.1]
  def change
    create_table(:horses) do |t|
      t.column :content, :text
      t.column :remind_at, :datetime
    end
  end
end

This migration will create the horses table for you on the way up, and automatically figure out how to drop the table on the way down.

Some commands cannot be reversed. If you care to define how to move up and down in these cases, you should define the up and down methods as before.

If a command cannot be reversed, an ActiveRecord::IrreversibleMigration exception will be raised when the migration is moving down.

For a list of commands that are reversible, please see ActiveRecord::Migration::CommandRecorder.

Transactional Migrations

If the database adapter supports DDL transactions, all migrations will automatically be wrapped in a transaction. There are queries that you can’t execute inside a transaction though, and for these situations you can turn the automatic transactions off.

class ChangeEnum < ActiveRecord::Migration[7.1]
  disable_ddl_transaction!

  def up
    execute "ALTER TYPE model_size ADD VALUE 'new_value'"
  end
end

Remember that you can still open your own transactions, even if you are in a Migration with self.disable_ddl_transaction!.

Namespace

Module

Class

Methods

Attributes

[RW] name
[RW] version

Class Public methods

[](version)

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 611
    def self.[](version)
      Compatibility.find(version)
    end
🔎 See on GitHub

check_all_pending!()

Raises ActiveRecord::PendingMigrationError error if any migrations are pending for all database configurations in an environment.

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 696
      def check_all_pending!
        pending_migrations = []

        ActiveRecord::Tasks::DatabaseTasks.with_temporary_connection_for_each(env: env) do |connection|
          if pending = connection.migration_context.open.pending_migrations
            pending_migrations << pending
          end
        end

        migrations = pending_migrations.flatten

        if migrations.any?
          raise ActiveRecord::PendingMigrationError.new(pending_migrations: migrations)
        end
      end
🔎 See on GitHub

check_pending!(connection = ActiveRecord::Tasks::DatabaseTasks.migration_connection)

Raises ActiveRecord::PendingMigrationError error if any migrations are pending.

This is deprecated in favor of check_all_pending!

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 680
      def check_pending!(connection = ActiveRecord::Tasks::DatabaseTasks.migration_connection)
        ActiveRecord.deprecator.warn(<<-MSG.squish)
          The `check_pending!` method is deprecated in favor of `check_all_pending!`. The
          new implementation will loop through all available database configurations and find
          pending migrations. The prior implementation did not permit this.
        MSG

        pending_migrations = connection.migration_context.open.pending_migrations

        if pending_migrations.any?
          raise ActiveRecord::PendingMigrationError.new(pending_migrations: pending_migrations)
        end
      end
🔎 See on GitHub

current_version()

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 615
    def self.current_version
      ActiveRecord::VERSION::STRING.to_f
    end
🔎 See on GitHub

disable_ddl_transaction!()

Disable the transaction wrapping this migration. You can still create your own transactions even after calling disable_ddl_transaction!

For more details read the “Transactional Migrations” section above.

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 745
      def disable_ddl_transaction!
        @disable_ddl_transaction = true
      end
🔎 See on GitHub

load_schema_if_pending!()

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 712
      def load_schema_if_pending!
        if any_schema_needs_update?
          # Roundtrip to Rake to allow plugins to hook into database initialization.
          root = defined?(ENGINE_ROOT) ? ENGINE_ROOT : Rails.root

          FileUtils.cd(root) do
            Base.connection_handler.clear_all_connections!(:all)
            system("bin/rails db:test:prepare")
          end
        end

        check_pending_migrations
      end
🔎 See on GitHub

migrate(direction)

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 737
      def migrate(direction)
        new.migrate direction
      end
🔎 See on GitHub

new(name = self.class.name, version = nil)

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 792
    def initialize(name = self.class.name, version = nil)
      @name       = name
      @version    = version
      @connection = nil
    end
🔎 See on GitHub

Instance Public methods

announce(message)

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 996
    def announce(message)
      text = "#{version} #{name}: #{message}"
      length = [0, 75 - text.length].max
      write "== %s %s" % [text, "=" * length]
    end
🔎 See on GitHub

connection()

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 1027
    def connection
      @connection || ActiveRecord::Tasks::DatabaseTasks.migration_connection
    end
🔎 See on GitHub

copy(destination, sources, options = {})

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 1048
    def copy(destination, sources, options = {})
      copied = []

      FileUtils.mkdir_p(destination) unless File.exist?(destination)
      schema_migration = SchemaMigration::NullSchemaMigration.new
      internal_metadata = InternalMetadata::NullInternalMetadata.new

      destination_migrations = ActiveRecord::MigrationContext.new(destination, schema_migration, internal_metadata).migrations
      last = destination_migrations.last
      sources.each do |scope, path|
        source_migrations = ActiveRecord::MigrationContext.new(path, schema_migration, internal_metadata).migrations

        source_migrations.each do |migration|
          source = File.binread(migration.filename)
          inserted_comment = "# This migration comes from #{scope} (originally #{migration.version})\n"
          magic_comments = +""
          loop do
            # If we have a magic comment in the original migration,
            # insert our comment after the first newline(end of the magic comment line)
            # so the magic keep working.
            # Note that magic comments must be at the first line(except sh-bang).
            source.sub!(/\A(?:#.*\b(?:en)?coding:\s*\S+|#\s*frozen_string_literal:\s*(?:true|false)).*\n/) do |magic_comment|
              magic_comments << magic_comment; ""
            end || break
          end

          if !magic_comments.empty? && source.start_with?("\n")
            magic_comments << "\n"
            source = source[1..-1]
          end

          source = "#{magic_comments}#{inserted_comment}#{source}"

          if duplicate = destination_migrations.detect { |m| m.name == migration.name }
            if options[:on_skip] && duplicate.scope != scope.to_s
              options[:on_skip].call(scope, migration)
            end
            next
          end

          migration.version = next_migration_number(last ? last.version + 1 : 0).to_i
          new_path = File.join(destination, "#{migration.version}_#{migration.name.underscore}.#{scope}.rb")
          old_path, migration.filename = migration.filename, new_path
          last = migration

          File.binwrite(migration.filename, source)
          copied << migration
          options[:on_copy].call(scope, migration, old_path) if options[:on_copy]
          destination_migrations << migration
        end
      end

      copied
    end
🔎 See on GitHub

down()

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 948
    def down
      self.class.delegate = self
      return unless self.class.respond_to?(:down)
      self.class.down
    end
🔎 See on GitHub

exec_migration(conn, direction)

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 976
    def exec_migration(conn, direction)
      @connection = conn
      if respond_to?(:change)
        if direction == :down
          revert { change }
        else
          change
        end
      else
        public_send(direction)
      end
    ensure
      @connection = nil
      @execution_strategy = nil
    end
🔎 See on GitHub

execution_strategy()

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 798
    def execution_strategy
      @execution_strategy ||= ActiveRecord.migration_strategy.new(self)
    end
🔎 See on GitHub

method_missing(method, *arguments, &block)

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 1031
    def method_missing(method, *arguments, &block)
      say_with_time "#{method}(#{format_arguments(arguments)})" do
        unless connection.respond_to? :revert
          unless arguments.empty? || [:execute, :enable_extension, :disable_extension].include?(method)
            arguments[0] = proper_table_name(arguments.first, table_name_options)
            if method == :rename_table ||
              (method == :remove_foreign_key && !arguments.second.is_a?(Hash))
              arguments[1] = proper_table_name(arguments.second, table_name_options)
            end
          end
        end
        return super unless execution_strategy.respond_to?(method)
        execution_strategy.send(method, *arguments, &block)
      end
    end
🔎 See on GitHub

migrate(direction)

Execute this migration in the named direction

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 955
    def migrate(direction)
      return unless respond_to?(direction)

      case direction
      when :up   then announce "migrating"
      when :down then announce "reverting"
      end

      time = nil
      ActiveRecord::Tasks::DatabaseTasks.migration_connection.pool.with_connection do |conn|
        time = Benchmark.measure do
          exec_migration(conn, direction)
        end
      end

      case direction
      when :up   then announce "migrated (%.4fs)" % time.real; write
      when :down then announce "reverted (%.4fs)" % time.real; write
      end
    end
🔎 See on GitHub

next_migration_number(number)

Determines the version number of the next migration.

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 1115
    def next_migration_number(number)
      if ActiveRecord.timestamped_migrations
        [Time.now.utc.strftime("%Y%m%d%H%M%S"), "%.14d" % number].max
      else
        "%.3d" % number.to_i
      end
    end
🔎 See on GitHub

proper_table_name(name, options = {})

Finds the correct table name given an Active Record object. Uses the Active Record object’s own table_name, or pre/suffix from the options passed in.

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 1106
    def proper_table_name(name, options = {})
      if name.respond_to? :table_name
        name.table_name
      else
        "#{options[:table_name_prefix]}#{name}#{options[:table_name_suffix]}"
      end
    end
🔎 See on GitHub

reversible()

Used to specify an operation that can be run in one direction or another. Call the methods up and down of the yielded object to run a block only in one given direction. The whole block will be called in the right order within the migration.

In the following example, the looping on users will always be done when the three columns ‘first_name’, ‘last_name’ and ‘full_name’ exist, even when migrating down:

class SplitNameMigration < ActiveRecord::Migration[7.1]
  def change
    add_column :users, :first_name, :string
    add_column :users, :last_name, :string

    reversible do |dir|
      User.reset_column_information
      User.all.each do |u|
        dir.up   { u.first_name, u.last_name = u.full_name.split(' ') }
        dir.down { u.full_name = "#{u.first_name} #{u.last_name}" }
        u.save
      end
    end

    revert { add_column :users, :full_name, :string }
  end
end
📝 Source code
# File activerecord/lib/active_record/migration.rb, line 900
    def reversible
      helper = ReversibleBlockHelper.new(reverting?)
      execute_block { yield helper }
    end
🔎 See on GitHub

revert(*migration_classes, &block)

Reverses the migration commands for the given block and the given migrations.

The following migration will remove the table ‘horses’ and create the table ‘apples’ on the way up, and the reverse on the way down.

class FixTLMigration < ActiveRecord::Migration[7.1]
  def change
    revert do
      create_table(:horses) do |t|
        t.text :content
        t.datetime :remind_at
      end
    end
    create_table(:apples) do |t|
      t.string :variety
    end
  end
end

Or equivalently, if TenderloveMigration is defined as in the documentation for Migration:

require_relative "20121212123456_tenderlove_migration"

class FixupTLMigration < ActiveRecord::Migration[7.1]
  def change
    revert TenderloveMigration

    create_table(:apples) do |t|
      t.string :variety
    end
  end
end

This command can be nested.

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 843
    def revert(*migration_classes, &block)
      run(*migration_classes.reverse, revert: true) unless migration_classes.empty?
      if block_given?
        if connection.respond_to? :revert
          connection.revert(&block)
        else
          recorder = command_recorder
          @connection = recorder
          suppress_messages do
            connection.revert(&block)
          end
          @connection = recorder.delegate
          recorder.replay(self)
        end
      end
    end
🔎 See on GitHub

reverting?()

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 860
    def reverting?
      connection.respond_to?(:reverting) && connection.reverting
    end
🔎 See on GitHub

run(*migration_classes)

Runs the given migration classes. Last argument can specify options:

  • :direction - Default is :up.

  • :revert - Default is false.

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 928
    def run(*migration_classes)
      opts = migration_classes.extract_options!
      dir = opts[:direction] || :up
      dir = (dir == :down ? :up : :down) if opts[:revert]
      if reverting?
        # If in revert and going :up, say, we want to execute :down without reverting, so
        revert { run(*migration_classes, direction: dir, revert: true) }
      else
        migration_classes.each do |migration_class|
          migration_class.new.exec_migration(connection, dir)
        end
      end
    end
🔎 See on GitHub

say(message, subitem = false)

Takes a message argument and outputs it as is. A second boolean argument can be passed to specify whether to indent or not.

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 1004
    def say(message, subitem = false)
      write "#{subitem ? "   ->" : "--"} #{message}"
    end
🔎 See on GitHub

say_with_time(message)

Outputs text along with how long it took to run its block. If the block returns an integer it assumes it is the number of rows affected.

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 1010
    def say_with_time(message)
      say(message)
      result = nil
      time = Benchmark.measure { result = yield }
      say "%.4fs" % time.real, :subitem
      say("#{result} rows", :subitem) if result.is_a?(Integer)
      result
    end
🔎 See on GitHub

suppress_messages()

Takes a block as an argument and suppresses any output generated by the block.

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 1020
    def suppress_messages
      save, self.verbose = verbose, false
      yield
    ensure
      self.verbose = save
    end
🔎 See on GitHub

up()

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 942
    def up
      self.class.delegate = self
      return unless self.class.respond_to?(:up)
      self.class.up
    end
🔎 See on GitHub

up_only(&block)

Used to specify an operation that is only run when migrating up (for example, populating a new column with its initial values).

In the following example, the new column published will be given the value true for all existing records.

class AddPublishedToPosts < ActiveRecord::Migration[7.1]
  def change
    add_column :posts, :published, :boolean, default: false
    up_only do
      execute "update posts set published = 'true'"
    end
  end
end
📝 Source code
# File activerecord/lib/active_record/migration.rb, line 919
    def up_only(&block)
      execute_block(&block) unless reverting?
    end
🔎 See on GitHub

write(text = "")

📝 Source code
# File activerecord/lib/active_record/migration.rb, line 992
    def write(text = "")
      puts(text) if verbose
    end
🔎 See on GitHub