Create database changes

Create is the developer's stage. It covers everything needed to author a governed database change, from a new project to a committed changeset.

The first pass runs in order. Review your scope and infrastructure, start your project with init, create and organize your changelogs, validate and preview the change, connect your database, secure your credentials, and define rollback. Commit the change when it is ready for review.

Most of the connection and credential work is one-time setup for each application. Authoring, validating, and committing happen with every change, and new changelog files appear with every release. A new application needs a new root changelog and a new registration, which is where Register picks up.

In this section

• Review your scope and infrastructure

• Use init to start your project

• Create a database change

• Validate and preview a change

• Commit the change

Create and organize changelogs

Structure your changelog, choose formats and organization, and manage changesets and attributes.

• How should I structure my changelog?

• Use SQL or modeled changelogs

• Changelog examples: SQL changelog example

• Use developer, IDE, and AI authoring tools

• Create your changelog

• Create your changelog from an existing database

Connect your database

Configure connection settings and find your database in the catalog.

• Configure connection settings

• What databases are supported by Liquibase?

• Databases included in the Liquibase AWS Marketplace image

• Using Liquibase and AWS RDS

• AlloyDB: What support does Liquibase have for AlloyDB for PostgreSQL?

Secure your credentials

Keep credentials out of files with environment variables, pipeline secrets, and vault extensions.

• Secure setup using environment variables

• Secure setup using GitHub actions secrets

• Secure setup using Azure key Vault

• Use the Liquibase HashiCorp vault extension

• Use Azure Key Vault secrets with Liquibase

• Use Google Secret Manager secrets with Liquibase

Store changelogs and outputs remotely

Keep changelogs, snapshots, and outputs in S3 or Azure Blob Storage.

• What are the rules for remote file paths?

• What are the rules for using search path with remote files?

• What Liquibase commands are used for remote files?

• Amazon S3: Connect Liquibase to Amazon S3

• Azure Blob Storage: Connect to Azure blob storage

Define rollback

Author rollback statements and choose a rollback strategy.

• Roll back a change

• Create custom rollback statements in Liquibase

• Implement a rollback strategy: Implement a rollback strategy with SQL changelogs

• What is a rollback?

• What automatic rollbacks does Liquibase support?

• What are targeted rollbacks?