Skip to content

README.md

Latest commit

 

History

History
81 lines (51 loc) · 3.76 KB

README.md

File metadata and controls

81 lines (51 loc) · 3.76 KB

How to use Flyway

Documentation Link

Running Flyway

Use the ✳️ Run flyway command on redshift workflow on GitHub Actions.

Inputs

Use workflow from

The branch to run the workflow on (i.e. the branch whose Flyway migrations you want applied). This will probably be a feature branch you are working on, or main.

AWS environment

The AWS environment to run the workflow on.

Flyway command to run

Flyway command to run. For more on the different commands, refer to the documentation here.

Database on which to perform migrations

Name of the redshift database you wish to run migrations on. There must be a corresponding subfolder in redshift-scripts/flyway/migrations or the job will fail. See section below under the Repository Files heading for more on this.

Repository Files

The files for Flyway are under the redshift-scripts/flyway/ directory.

lib/

Library files for Flyway (library itself and the Redshift JAR). These can be left alone for the most part unless you need to update one or the other.

migrations/

Migration files for Flyway. For more on the concept, refer to the documentation here.

flyway.conf

Config file for flyway. We set some properties in the lambda as environment variables, but set the following static properties in the flyway.conf file. For more on the config file see the documentation here

  • flyway.defaultSchema - the schema where flyway will put its history table. we give it its own schema to do this
  • flyway.schemas - list of schemas flyway manages. it will try and create them if they don't exist in an empty database, and it will clean them all when the clean command is run. The migrate command will operate from the default schema so all schema references must be explicit. If we didn't set a default schema, flyway would use the first schema in this list as the default one

See the documentation here for more on these schema options

Database subdirectories

Note that all migration files are under subdirectories of the migrations/ directory. These subdirectories are the names of databases in Redshift. This means that

  • Each database has a separate set of migrations and version numbers do not conflict between them (e.g. two different databases may have a V1.2 migration)
  • The database chosen for the GitHub workflow must be one of these and will determine which migrations are run and on which database
An example

Consider the following directory layout

redshift-scripts/
  flyway/
    database_one/
      V1__create_user_table.sql
      V2__add_user_email_column.sql
    database_two/
      V1__set_up_schemas.sql
  • Running the workflow with a database input of database_one will run V1__create_user_table.sql and V2__add_user_email_column.sql on the database_one database in redshift
    • The migrations will run with reference to "database_one"."flyway_schema_history";
    • No other migrations will be run (if you also wanted to migrate database_two you would need to run the workflow a second time with that as the database input)
    • All other databases will be unaffected

Migration types

Flyway allows you to use versioned or repeatable migrations. For descriptions of both types, refer to the documentation here.