bigquery-migrations 0.5.3

Simple tool for writing Google BigQuery migrations

python-bigquery-migrations

The python-bigquery-migrations package provides a streamlined way to create and manage BigQuery databases using intuitive CLI commands, such as the following:

bigquery-migrations run

What are the benefits of using migrations?

Migrations are like version control for your database, allowing you to define and share the application's datasets and table schema definitions.

Getting Started

0. Prerequisite

• Google Cloud Project with enabled billing
• Enabled Google Cloud BigQuery API
• Google Cloud Service Account JSON file

1. Install

pip install bigquery-migrations

2. Create the project folder structure

Create two subdirectory:

1. credentials
2. migrations

your-project-root-folder
├── credentials
├── migrations
└── ...

3. Create the neccessary files in the folders

Google Cloud Service Account JSON file

Put your Google Cloud Service Account JSON file in the credentials subdirectory. See more info in the Authorize BigQuery Client section

your-project
├── credentials
│   ├── gcp-sa.json
├── migrations
└── ...

You can use different folder name and file name but in that case you must specify them with command arguments, such as the following:

bigquery-migrations run --gcp-sa-json-dir  my-creds --gcp-sa-json-fname my-service-account.json

argument	description
--gcp-sa-json-dir	Name of the service account JSON file directory (optional)
--gcp-sa-json-fname	Name of the service account JSON file (optional)

IMPORTANT!
Never check the Google Cloud Service Account JSON file into version control. This file contains sensitive credentials that could compromise your Google Cloud account if exposed.

To prevent accidental commits, make sure to add the file to your .gitignore configuration. For example:

# .gitignore
gcp-sa.json

By ignoring this file, you reduce the risk of unintentional leaks and maintain secure practices in your repository.

Migrations

Create your own migrations and put them in the migrations directory. See the Migration structure section and Migration naming conventions section for more info.

your-project
├── credentials
│   ├── gcp-sa.json
├── migrations
│   ├── 2024_12_01_120000_create_users_table.py
└── ...

You can use different folder name but in that case you must specify it with a command argument:

bigquery-migrations run --migrations-dir bq-migrations

argument	description
--migrations-dir	Name of the migrations directory (optional)

Running migrations

IMPORTANT!
You have to create your own Migrations first! Jump to Creating Migrations section

To run all of your outstanding migrations, execute the run command:

bigquery-migrations run

You can specify the Google Cloud Project id with the --gcp-project-id argument:

bigquery-migrations run --gcp-project-id your-gcp-id

Migration log

IMPORTANT!
It's cruical to keep the migration_log.json file in place, and not to modify it manualy!

After the first successful run a migration_log.json is created in the migrations directory.

your-project
├── migrations
│   ├── 2024_12_01_120000_create_users_table.py
    ├── migration_log.json
...

The migration_log.json file content should look like this:

{
    "last_migration": "2024_12_10_121000_create_users_table",
    "timestamp": "2024-12-18T12:25:54.318426+00:00"
}

Rolling Back Migrations

Rollback the last migration

To reverse the last migration, execute the rollback command and pass last with the --migration-name argument:

bigquery-migrations rollback --migration-name last

Rollback a specific migration

To reverse a specific migration, execute the rollback command and pass the migration name with the --migration-name argument:

bigquery-migrations rollback --migration-name 2024_12_10_121000_create_users_table

Rollback all migrations

To reverse all of your migrations, execute the reset command:

bigquery-migrations reset

Authorize BigQuery Client

Put your service account JSON file in the credentials subdirectory in the root of your project.

your-project
├── credentials
│   ├── gcp-sa.json
...

Creating a Service Account for Google BigQuery

You can connect to BigQuery with a user account or a service account. A service account is a special kind of account designed to be used by applications or compute workloads, rather than a person. Service accounts don’t have passwords and use a unique email address for identification.

To create a BigQuery service account key

1. Sign in to the Google Cloud management console.
2. Make sure that you have API enabled on your BigQuery API page. If you don’t see API Enabled, choose Enable.
3. On the Service accounts page, choose your BigQuery project, and then choose Create service account.
4. On the Service account details page, enter a descriptive value for Service account name. Choose Create and continue. The Grant this service account access to the project page opens.
5. For Select a role, choose BigQuery, and then choose BigQuery Admin.
6. Choose Continue, and then choose Done.
7. On the Service account page, choose the service account that you created.
8. Choose Keys, Add key, Create new key.
9. Choose JSON, and then choose Create. Choose the folder to save your private key or check the default folder for downloads in your browser.

Creating migrations

Put your migrations files in the migrations subdirectory of the root of your project.

your-project
├── migrations
│   ├── 2024_12_01_120000_create_users_table.py
...

Migration structure

The migration class must contain two methods: up and down.

The up method is used to add new dataset, tables, columns etc. to your BigQuery project, while the down method should reverse the operations performed by the up method.

from google.cloud import bigquery
from bigquery_migrations import Migration

class CreateUsersTable(Migration):
    """
    See:
    https://github.com/googleapis/python-bigquery/tree/main/samples
    """

    def up(self):
        # TODO: Set table_id to the ID of the table to create.
        table_id = "your_project.your_dataset.example_table"
        
        # TODO: Define table schema
        schema = [
            bigquery.SchemaField("id", "INTEGER", mode="REQUIRED"),
            bigquery.SchemaField("name", "STRING", mode="REQUIRED"),
            bigquery.SchemaField("created_at", "TIMESTAMP", mode="NULLABLE"),
        ]
        table = bigquery.Table(table_id, schema=schema)
        table = self.client.create_table(table)
        print(
            "Created table {}.{}.{}".format(table.project, table.dataset_id, table.table_id)
        )

    def down(self):
        # TODO: Set table_id to the ID of the table to fetch.
        table_id = "your_project.your_dataset.example_table"
        
        # If the table does not exist, delete_table raises
        # google.api_core.exceptions.NotFound unless not_found_ok is True.
        self.client.delete_table(table_id, not_found_ok=True)
        print("Deleted table '{}'.".format(table_id))

Migration naming conventions

Pattern	yyyy_mm_dd_hhmmss_your_class_name.py
Example filename	2024_12_10_120000_create_users_table.py
Example class name	CreateUsersTable

Changelog

0.5.3

Security

• Upgraded protobuf from 5.29.1 to 6.33.6 to fix two high-severity CVEs:
  • JSON recursion depth bypass (protobuf < 5.29.6)
  • Denial of Service via recursive groups in pure-Python backend (protobuf < 5.29.5)
• Upgraded google-cloud-bigquery from 3.27.0 to 3.40.1
• Upgraded google-auth from 2.37.0 to 2.49.1
• Upgraded google-api-core from 2.24.0 to 2.30.0
• Upgraded grpcio / grpcio-status from 1.68.1 to 1.78.0
• Added explicit protobuf>=5.29.6 lower-bound in pyproject.toml to protect downstream consumers

Fix

• migration_cli.py: --gcp-sa-json-dir and --gcp-sa-json-fname CLI arguments were silently ignored due to both reading from args.migrations_dir instead of their own argument attributes

Removed

• Dropped Python 3.9 support (EOL since October 2025). The updated security dependency chain (cryptography>=46, cffi>=2.0, pycparser==3.0) requires Python >=3.10. Minimum supported version is now Python 3.10

0.5.2.

Feature

• Rollback the last migration

Documentation

• README.md
  • Modified sections
    • Getting Started
    • Rolling Back Migrations

0.5.1

Documentation

• README.md
  • Modified sections
    • Create the neccessary files in the folders

0.5.0

Feature

• Rollback to a specific migration

Documentation

• README.md
  • New sections
    • Migration log
  • Modified sections
    • Running migrations
    • Rollback migrations

0.4.3

Documentation

• README.md
  • GCP Service account creation process updated

0.4.2

Documentation

• README.md
  • Sample code: import correction
  • New sections:
    • GCP Service account creation process
    • Migration naming convention

0.4.1

Documentation

• README.md sample code: removed unnecessary lines of code

0.4.0

This is the first release which uses the CHANGELOG file.