We currently recommend that you incorporate seed scripts into your images. Another way to seed your environments is to create a separate service on your compose file.

​

Building images with seeds

In most cases, using a built image with a seed script is the easiest way to seed a preview environment.

​

Mongo example

If you’re using MongoDB, you can create a Dockerfile using the mongo image as the base image, and copy the necessary seed files to the /docker-entrypoint-initdb.d path within that image. Any files within that path will be automatically executed by the image when the mongo service starts.

Assume you have the seed script below, named seed.js within a scripts folder, for example.

db.test.drop();
db.test.insertMany([
    {
        _id: 1,
        name: "Alice",
        age: 50,
    },
    {
        _id: 2,
        name: "Bob",
        age: 75,
    },
]);

In that case, you’ll simply configure your Dockerfile so that the JavaScript file is copied to /docker-entrypoint-initdb.d when building the image.

FROM mongo:3.6

COPY ./scripts/seed.js /docker-entrypoint-initdb.d

Finally, you’ll reference this Dockerfile in the docker-compose.yml file used by Ergomake.

version: "3"

services:
    db:
        build:
            context: ../mongo
            dockerfile: Dockerfile
        restart: always
        environment:
            MONGO_INITDB_ROOT_USERNAME: { { USERNAME } }
            MONGO_INITDB_ROOT_PASSWORD: { { PWD } }
            MONGO_INITDB_DATABASE: { { DBNAME } }

.
├──── .ergomake
│   └── docker-compose.yml <--- Used by Ergomake
└──── mongo
    ├── Dockerfile
    └── scripts
        └── seed.js

​

Postgres Example

If you’re using the postgres image, you can simply copy your seed scripts to /docker-entrypoint-initb.d. That’s because the image is configured to execute any files within that directory when the postgres service starts.

Assume you had the seed script below, named dump.sql, within a scripts folder.

CREATE TABLE IF NOT EXISTS users (
  "id" TEXT NOT NULL,
  "name" TEXT NOT NULL,
  "email" TEXT NOT NULL,

  PRIMARY KEY ("id")
);

INSERT INTO users (id, name, email) VALUES ('a123', 'My Name', 'name@example.com');

In that case, you’d first create a script to run it, let’s say seed.sh, within that same scripts folder.

#!/bin/bash

psql -U $POSTGRES_USER -d $POSTGRES_DB -a -f /app/scripts/db/dump.sql

Then, you’d configure your Dockerfile to copy seed.sh the dump.sql to docker-entrypoint-initdb.d, as shown below.

FROM postgres:12
COPY ./scripts/init.sh /docker-entrypoint-initdb.d
COPY ./scripts/dump.sql ./scripts/db/dump.sql

Finally, you’ll reference that Dockerfile in the docker-compose.yml you use for Ergomake.

services:
    db:
        build:
            context: ../postgres
            dockerfile: Dockerfile
        environment:
            - POSTGRES_USER=user
            - POSTGRES_PASSWORD=password
            - POSTGRES_DB=database
        ports:
            - 5432:5432

.
├──── .ergomake
│   └── docker-compose.yml <--- Used by Ergomake
└──── postgres
    ├── Dockerfile
    └── scripts
        ├── dump.sql
        └── seed.sh

​

Seeding environments with services

To seed an environment with a separate service, all you need to do is create a separate service in your compose and change its command so that it runs the seed command.

version: '3.8'
services:
  api:
    build: ../backend
    ports:
      - '3001:3001'

  seed:
    build: ../backend
    # Instead of running your image's command, you can use this one to override it
    command: ./my-seed-script.sh
    ports:
      - '3001:3001'

Another way to seed it and still have a single service is to use multiple commands on your compose.

version: '3.8'
services:
  api:
    build: ../backend
    # Here you'll run both your seed and your app
    command: bash -c "python manage.py migrate && python manage.py runserver 0.0.0.0:8000"
    ports:
      - '3001:3001'