Skip to content

Test a Node server with a Postgres database with docker and docker compose

Notifications You must be signed in to change notification settings

talagbada/test-node-with-docker

 
 

Repository files navigation

Test a Node.js server with a database with Docker and docker-compose

This repository contains sample code on how to a Node.js app with Docker and docker-compose.

Please help this repo with a ⭐️ if you find it useful! 😁

This repository is part of the Test Node with Docker tutorial series provided by productioncoder.com.

Test Node with Docker

For updates, please reach out to @productioncoder on Twitter.

2. How to run this application

2.1. Running the project locally

You can run the project locally with your own local Postgres instance

npm run dev

Make sure to update the configuration app with your database credentials in the .config/local.json file.

2.2. Running with docker compose

2.2.1 docker-compose up

Running:

docker-compose up

will spin up:

  1. express server itself (internally port 8080) forwarded to port 8080 on the host machine
  2. Postgres database (internally 5432), forwarded to port 2345

We forward to port 2345 on purpose so it does not collide with a local Postgres instance that might be running on the host machine.

To run in detached mode, use

docker-compose up -d

2.2.2 clean docker-compose run

If you would like to do a clean run (i.e. remove all data in already stopped containers), you can run

docker container prune

to remove all unused containers

As an alternative you can run

docker system prune

to clean up all unused resources. Note that this removes all the unused images, volumens etc. on your host machine.

2.2.3 docker-compose stop vs docker-compose down

Note that typing ctrl + c when running docker-compose up will only stop the container, but not remove it - which is equivalent of running:

docker-compose stop

I.e. if you start up the container again, the database will contain the data you have put in before.

Please refer to the above section on the clean docker-compose section on how to clean up all unused containers.

2.2. Running the webserver with docker but accessing db on host machine

If you want to to only run the webapp in a docker container, you need to adapt your configuration first.

Since the docker container itself is isolated, setting the database host to localhost would result in the container trying to connect to itself to connect to the database.

To connect to the database running on your host machine, you need to set the host name to host.docker.internal.

Change the hostname field under the db object inside the ./config.local.json like so.

{
  "db": {
    "host": "host.docker.internal"
  }
}

You can leave the remaining config as it was.

This is a special DNS entry within the container that allows the container to connect to your host machine. Please refer to the documentation for more details.

Once you have updated your config, you need to build an image and start a container from it.

Execute the following command from the root directory of the project (where the Dockerfile is located)

docker image build -t productioncoder/test-node-with-docker:local .

After that you can start a container

docker container run -p 8080:8080 productioncoder test-node-with-docker:local

The -p flag publishes the port from the docker container to your host machine.

3 Testing

3.1. Testing with docker-compose

Running

npm test

will spin up a Postgres database (internally port 5432), port-forwarded to port 2345 on the host machine.

It will then run all the tests with mocha as a test runner and then destroy the containers afterwards.

By testing with a real database, we achieve a high test-fidelity.

3.2 Running tests locally

You can run the tests with your local Postgres database by running:

npm run test-local

Note that this add data to your local database and that the tests might fail if unique constraints are violated.

About

Test a Node server with a Postgres database with docker and docker compose

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • JavaScript 83.8%
  • Shell 13.4%
  • Dockerfile 2.8%