simple-db-migrate “quick” documentation

Quick start

simple-db-migrate is damn simple. The best way to understand how it works is by installing and using it.

You can install it by typing:

    $ sudo easy_install simple-db-migrate

There are more detailed instructions and information about installing directly from source on the project website.

After installing, for usage tips type:

    $ db-migrate --help

Upgrading simple-db-migrate version

The 1.5.0 version removed some legacy code which updates the version table from the old format to the actual.
To whom use simple-db-migrate for MySQL or Oracle with version less than 1.4.1.1, and the Bruno Caimar version for MS-SQL Server should use the version 1.4.4 before use the new version, to have the version table updated.

Understanding how it works

The first thing you’ll need is a migration file. There are some example migration files in the “example” directory. The migration files have the following format:

    SQL_UP = """
    CREATE TABLE aleatory (
      id int(11) NOT NULL auto_increment,
      name varchar(255) default NULL,
      PRIMARY KEY  (id)
    );
    """
    SQL_DOWN = """
    DROP TABLE aleatory;
    """

... where SQL_UP and SQL_DOWN are two strings that contains respectively the SQL statements to upgrade and downgrade the database schema.

You can use db-migrate to create new migrations by typing:

    $ db-migrate --create create_table_users

The file names need to respect the format “YYYYMMDDHHMMSS_migration_description.migration”. simple-db-migrate uses the YYYYMMDDHHMMSS information to track the database schema version and to decide the order of execution of the scripts. simple-db-migrate will go through all .migration files in your directory and execute all of them in their creation (date) order.

Second, you have to configure access to your database so simple-db-migrate can execute DDL. Just create a file named “simple-db-migrate.conf”, with the following content (there is also an example in the “example” directory):

    DATABASE_HOST = "localhost"
    DATABASE_USER = "root"
    DATABASE_PASSWORD = ""
    DATABASE_NAME = "migration_example"
    DATABASE_MIGRATIONS_DIR = "."

*** The MIGRATIONS_DIR directive may have relative (from the location of this file) or absolute paths separated by _:_ pointing to the migrations directories.

You don’t need to create the database. simple-db-migrate will create it for you.

After these two things you are ready to go. Just navigate to the directory where you created your configuration file and type:

    $ db-migrate

If you don’t want to navigate to the directory, you can specify it path instead. In this case you will also need to specify the path to the config file. Note that this also makes it possible to use any name you like for the config file:

    $ db-migrate --config=path/to/file.conf

Migrating to a specific version

If you want, you can migrate your database schema to a specific version by supplying the —migration (or -m) parameter. The version id is the YYYYMMDDHHMMSS identifier used at the migration file:

    $ db-migrate --migration=20090227000129

If you don’t specify any version, simple-db-migrate will migrate the schema to the latest version available in the migrations directories specified in the config file.

Configuring multiple database environments

If you want to use the same configuration file for multiple environments you can prefix the names with the name of your environment and specify it when executing, like the example below.

    DATABASE_HOST = "localhost"                              # default database host
    STAGING_DATABASE_HOST = "staging.host.com"               # staging database host
    PRODUCTION_DATABASE_HOST = "production.host.com"         # production database host
    DATABASE_USER = "root"
    DATABASE_PASSWORD = ""
    DATABASE_NAME = "migration_example"
    DATABASE_MIGRATIONS_DIR = "."

    $ db-migrate --config=path/to/file.conf                  # will use default configurations
    $ db-migrate --config=path/to/file.conf --env=staging    # will use staging configurations, and default to keys not prefixed
    $ db-migrate --config=path/to/file.conf --env=production # will use production configurations, and default to keys not prefixed

Available configurations

You can set default values for internal configurations in your configuration file and overwrite (some of them) using the command line parameters. Below is a list of all configuration options.

Configuration	description	default value	possible values
DATABASE_HOST	hostname where database is located	–	–
DATABASE_USER	username used to connect to database and execute the commands	–	–
DATABASE_PASSWORD	password used to connect to database and execute the commands	–	–
DATABASE_NAME	database name used where the commands will be executed	–	–
DATABASE_ENGINE	the database type where migrations will be executed	mysql	oracle,mysql,mssql
DATABASE_VERSION_TABLE	the table name used to save database versions	db_version	any name supported by the database
UTC_TIMESTAMP	create migration files using UTC time to format the name	False	True,False
DATABASE_MIGRATIONS_DIR	directories to look for migration files separated by :	–	–
DATABASE_ENCODING	encoding used on database	utf-8	any valid encoding
DATABASE_SCRIPT_ENCODING	encoding used on migration files	utf-8	any valid encoding
drop_db_first	if True drop the database before executing migrations	False	True,False

force_execute_old_migrations_versions	if True and current and destination database versions are equal, execute any old migrations not executed yet	False	True,False
force_use_files_on_down	if True use SQL_DOWN from migration files instead of that present on version table	False	True,False
label_version	label to be applied to all executed migrations when doing a upgrade on database	–	–
log_dir	directory where a file will be created with a full log of the process, with the current time as name	–	–
new_migration	name for the migration to be created	–	any alpha numeric word, without spaces
paused_mode	execute migrations, pausing after finishing each one	False	True,False
schema_version	the desired version of the database, will do a upgrade or a downgrade to be sure that this will be the current version of database	–	–
show_sql	if True show executed SQL commands	False	True,False
show_sql_only	if True only show the SQL, but do not execute them	False	True,False

Supported databases engines

You can use this project to run migrations on MySQL, Oracle and MS-SQL server databases.
The default database engine is MySQL. To use the other databases set the DATABASE_ENGINE constant in the configuration file.

Roadmap, bug reporting and feature requests

For detailed info about future versions, bug reporting and feature requests, go to issues page.

ruxkor/simple-db-migrate