Migration file format reference

Migration filename format

Each migration file has the following format:

<version>_<name>.{up|down}.{sql|yaml}

A version which is the Unix timestamp in nanoseconds when the file was created is the first part. Followed by a name which is either manually added or auto-generated by the console. The next part indicates what step this is. If it is up, it means that this is the forward step, for e.g. creating a relationship or a table. The down indicates that it is the corresponding rollback step. The last part denotes the file format - whether it is an SQL file or a YAML file.

YAML files

Each up or down YAML file contains Hasura Metadata API actions to be executed for that particular migration. The file should have a list of such API actions, which is represented with a - in YAML. Here is an example file:

1551151778747_create_table_public_author.up.yaml

- type: run_sql
  args:
    sql: |
      CREATE TABLE "public"."author" (
        "id" SERIAL NOT NULL,
        "name" TEXT NOT NULL,
        PRIMARY KEY ("id")
      );
- type: track_table
  args:
    schema: public
    name: author

1551151778747_create_table_public_author.down.yaml

- type: untrack_table
  args:
    schema: public
    name: author
- type: run_sql
  args:
    sql: |
      DROP TABLE "public"."author"

Each one of the actions in these files are actually Hasura Metadata API calls, which are executed in sequence when the migration is applied. You can find all the metadata actions that are available in the API reference.

SQL Files

Note

The files are there just for the user’s convenience, since the SQL can already be included in the YAML files.

As the name indicates, an SQL file can contain SQL statements which are executed on the apply step of the migration. There can be an up SQL and a down SQL file. The SQL file is executed first if it is present in an up migration and YAML file is executed first in a down migration.