Creating a Hasura Seed data Migration

Use case

It's often useful to add some initial Seed data into your database as part of the initialization process, with Hasura Seeds you can do that. This is particularly useful for adding a testing user or to pre-populate values if you have set a table in Hasura as an enum table to expose it as GraphQL enums in the GraphQL API.

Create a Hasura Seed file

You can use the Hasura CLI to automatically create Seed files for you based on the data which is already in a table in a database.

hasura seed create myAuthorsSeed --from-table author

Select the database containing the table you've specified.

? Select a database to use

CLI will log

INFO created seed file successfully     file=/Users/me/myDemoProject/seeds/default/1656499378904_myAuthorsSeed.sql

If you open the file created you will see standard pure SQL insert statements with the data contained in the database. Eg:

SET check_function_bodies = false;
INSERT INTO public.author (id, name) VALUES ('1', 'Woolf');
INSERT INTO public.author (id, name) VALUES ('2', 'Tolkien');
INSERT INTO public.author (id, name) VALUES ('3', 'Austen');
INSERT INTO public.author (id, name) VALUES ('4', 'Orwell');

Apply a Hasura Seed file

To apply a seeds file to your Hasura instance you can use the apply Hasura CLI command. Eg:

hasura seed apply --file 1656499378904_myNewAuthorsSeed.sql

If you do not specify a database with the --database-name flag then the CLI will prompt you to choose the database to apply the seeds to and as long as there are no conflicts and the SQL executes successfully to input the data then the CLI will log:

INFO Seeds planted

Success!