Initializing a MongoDB Docker Image with Data

Prerequisites

• Docker installed on your machine

Get Mongo Running via Docker Compose

First, create an empty directory to house a few files related to our task:

mkdir docker-mongo-import-example

Then create a new docker-compose.yml file with the following contents:

# docker-compose.yml
services:
  mongo:
    image: mongo:latest
    container_name: mongo
    ports:
      - "27017:27017"
    environment:
      MONGO_INITDB_ROOT_USERNAME: root
      MONGO_INITDB_ROOT_PASSWORD: example

  mongo-express:
    image: mongo-express:latest
    container_name: mongo-express
    ports:
      - "8081:8081"
    environment:
      ME_CONFIG_MONGODB_ADMINUSERNAME: root
      ME_CONFIG_MONGODB_ADMINPASSWORD: example
      ME_CONFIG_MONGODB_SERVER: mongo
      ME_CONFIG_BASICAUTH: false
    depends_on:
      - mongo

This will create a Docker Compose stack with two containers: MongoDB and mongo-express, which can be used to visualize our database in the browser.

Run the Docker Compose stack with docker compose up --build

Once everything is up and running, you should be able to use mongo-express to view the contents of the MongoDB database. Try navigating to http://localhost:8081 in your browser, and you should see a page like:

Try clicking around in the different databases that were created.

Not much to look at, huh? Let's fix that.

Download some test data

MongoDB has some links to sample data sets that can easily be imported into a MongoDB database: https://www.mongodb.com/resources/basics/databases/sample-database

For this example, I am using https://insideairbnb.com/get-the-data/. I chose a listings.csv.gz file for a city at random.

Download, unzip, and move the resulting listings.csv file to a new directory in our folder: import-data.

Add an import script

The MongoDB Docker image will run any .js or .sh files present in the /docker-entrypoint-initdb.d directory in the Docker container at startup. We will leverage this to help us easily import our sample dataset.

Add a new file in our project folder: init-files/load-data.sh

# init-files/load-data.sh
mongoimport --host 0.0.0.0 -c collection --type csv --file /import-data/listings.csv --headerline

This will use the mongoimport utility to connect to the MongoDB database (0.0.0.0 as the host since this is running inside the MongoDB Docker container).

Mount the volumes

With the sample data and import script in place, mount the volumes in the MongoDB Docker container so that MongoDB can utilize them during container startup.

# docker-compose.yml
services:
  mongo:
    image: mongo:latest
    container_name: mongo
    ports:
      - "27017:27017"
    environment:
      MONGO_INITDB_ROOT_USERNAME: root
      MONGO_INITDB_ROOT_PASSWORD: example
      MONGO_INITDB_DATABASE: listings
    # Add volumes section here
    volumes:
      # The following mount will be used for the import scripts
      - ./init-files:/docker-entrypoint-initdb.d
      # The following mount will be used for the import data
      - ./import-data:/import-data

  # ...

Now, with everything in place, run docker compose down --volumes (since we already initialized our MongoDB database earlier, we must remove the volume that was created) and then docker compose up --build.

Success! If everything worked properly, when you visit http://localhost:8081, you should see a new database, test. If you enter the database, then the collection, you should see data from your import file:

Importing JSON files

If you'd like to import a JSON file, you can use the following command:

mongoimport --collection collection --file /import-data/listings.json --jsonArray