Skip to content

Sqlalchemy - Alembic Migrations

SQLAlchemy: Alembic Migrations#

Alembic is a lightweight database migration tool for usage with the SQLAlchemy Database Toolkit for Python

Alembic - A chemical apparatus to purify substances by distillation

Basic Info#


Alembic uses SQLAlchemy and database drivers

python -m pip install alembic
  • Alembic supports Python versions 3.7 and above


Alembic provides for the creation, management, and invocation of change management scripts for a relational database, using SQLAlchemy as the underlying engine

Ideally you want it installed in your virtual environment so that when you run alembic it has access to your models.

The Migration Environment#

  • Starting point
  • A directory of scripts specific to the application
  • Created once and kept with applciation source code

To initialise and call the folder alembic:

alembic init alembic

Example directory:

  • alembic - this directory lives within your application’s source tree and is the home of the migration environment. It can be named anything, and a project that uses multiple databases may even have more than one.
  • - script run when the mgiration tool is invoked - mostly how to connect to the db
  • - the mako template used to generate migration scripts - creates the files with /version
  • versions/ - holds version scripts - numbering does not ascend - it uses guid that refer to each other. Versions from different branches can be spliced by hand

Alembic has other templates based on project setup:

alembic list_templates

Editing the Alembic.ini File#

Alembic placed a file alembic.ini into the current directory

The file is read using Python’s ConfigParser.SafeConfigParser object

  • file_template - this is the naming scheme used to generate new migration files. Uncomment the presented value if you would like the migration files to be prepended with date and time, so that they are listed in chronological order.
    • %%(rev)s - revision id
    • %%(slug)s - a truncated string derived from the revision message
    • %%(epoch)s - epoch timestamp based on the create date; this makes use of the Python datetime.timestamp() method to produce an epoch value.
    • %%(year)d, %%(month).2d, %%(day).2d, %%(hour).2d, %%(minute).2d, %%(second).2d - components of the create date, by default unless the timezone configuration option is also used.
  • sqlalchemy.url - A URL to connect to the database via SQLAlchemy.

For a single database, starting up all that is needed is:

sqlalchemy.url = postgresql://scott:tiger@localhost/test

Create a Migration Script#

alembic revision -m "create product table"

A new file is generated:

"""create product table

Revision ID: 5a5a17843e58
Create Date: 2022-09-15 16:39:28.057474

from alembic import op
import sqlalchemy as sa

# revision identifiers, used by Alembic.
revision = '5a5a17843e58'
down_revision = None
branch_labels = None
depends_on = None

def upgrade() -> None:

def downgrade() -> None:
  • Our job here is to populate the upgrade() and downgrade() functions with directives that will apply a set of changes to our database.
  • upgrade() is required while downgrade() is only needed if down-revision capability is desired

We add:

def upgrade():
        sa.Column('id', sa.Integer, primary_key=True),
        sa.Column('name', sa.String(50), nullable=False),
        sa.Column('description', sa.Unicode(200)),

def downgrade():

Run the Migration#

Run to the most recent migration:

alembic upgrade head

Run to a specific version:

alembic upgrade 5a5a17843e58
  • It checks if the alembic_version table exists - if not it creates it
  • It runs upgrade in each file until it reaches the given revision

Running the Second Migration#

alembic revision -m "Add a column"

And edit:

def upgrade() -> None:
    op.add_column('product', sa.Column('last_modified_date', sa.DateTime))

def downgrade() -> None:
    op.drop_column('product', 'last_modified_date')

Run the migration:

alembic upgrade 86f

A partial number can be used as long as the portion of the hash is unique

You can also give relative identifiers:

alembic upgrade +2 # move 2 versions up

alembic downgrade -1 # move 1 version down

alembic upgrade ae10+2 # move to ae10 + 2 versions

Getting Info#

Get the current hash:

alembic current

INFO  [alembic.runtime.migration] Context impl PostgresqlImpl.
INFO  [alembic.runtime.migration] Will assume transactional DDL.
86f384da7285 (head)

Get alembic history:

alembic history
5a5a17843e58 -> 86f384da7285 (head), Add a column
<base> -> 5a5a17843e58, create product table

Can also run it verbosely:

alembic history --verbose


To downgrade back to the beginning:

alembic downgrade base

Back up again:

alembic upgrade head

Auto Generating Migrations#

Alembic can view the status of the database and compare against the table metadata in the application, generating the “obvious” migrations based on a comparison.


alembic revision --autogenerate

This creates candidate migrations that can be reviewed manually

In you should add your model’s metadata object

# add your model's MetaData object here
# for 'autogenerate' support
# from myapp import mymodel
# target_metadata = mymodel.Base.metadata
target_metadata = None

changed to:

from myapp.mymodel import Base
target_metadata = Base.metadata

I tried the above and it did not work. I had to import the base model but then also import all the models afterwards which is made clear by this stackoverflow answer

This is sent into the run_migrations_online() function

Then we can run autogeneration - and it will check the MetaData against the database and create migraitons based on it.

alembic revision --autogenerate -m "Add existing tables"

Always look at the candidate migrations…first before applying.

alembic upgrade head

What does Autogenerate Detect (and not Detect)?#

Autogenerate is not perfect. One must always manually review the candidate migration.

Autogenerate will detect:

  • Table additions and removals
  • Column additions and removals
  • Change of nullable status on columns
  • Changes in indexes and named unique constraints
  • Changes in foreign key constraints

Autogenerate will optionally detect:

Autogenerate cannot detect:

  • Changes of table name - these will come out as an add/drop of two different tables, and should be hand-edited into a name change instead.
  • Changes of column name - these are detected as a column add/drop pair, which is not at all the same as a name change
  • Anonymously named constraints (remember to give constraints a name)
  • Special SQLAlchemy types such as Enum when generated on a backend which doesn’t support ENUM directly

Controlling what is Autogenerated#

Continue with Controlling what is Autogenerated

How to Start Afresh#

If you want to start from scratch again.

Remove all db tables and truncate the alembic_version table. Remove the migration files in versions/ Run the migration autogeneration if needed. Run the migrations.