Docker/Docker Compose Excercise

A small exercise that exercise that teaches the basic of Docker/Docker Compose as development tools. The repository contains an API that has one endpoint (/), that returns all the document from a MongoDB instance. The code for the API is under src. The repository also contains integration tests (under test). The test calls the single endpoint and makes sure it returns an empty array.

This guide was tested on:

• Windows - Docker version Version 2.0.0.0-win81 (29211)
• Mac - Docker version Version 2.0.0.3 (31259)

In case something is misbehave, make sure you have this version or above.

Step 0 - Running locally

Run the tests manually. To run the tests, we first need a MongoDB instance.

• Use docker to run a MongoDB instance on your machine. Make sure you use an official image. Hint: Mongo is running on port 27017, make sure you publish this port. To test connection use one of the following:
  • nc -vz localhost 27017 (Linux/OSX)
  • Test-NetConnection -Computer localhost -Port 27017 (Windows)
• Try to run the tests: yarn install && yarn test
• If you did everything correctly, the test will run successfully.

Step 1 - Adding a docker container

Local running is nice, but how can we run this in the CI? Let's start with writing a docker file for the API.

• Write a docker file for the API, and build it.
• Try to run the tests inside the docker file. Did it fail? How can we solve it?
• Change something in the code (for example, the server port), and re-build the docker image. Did yarn install command run? How can we make it more efficient?

Step 2 - Adding docker compose

Now that we have an image, let's add docker compose!

• Start with writing a basic docker-compose.yml file that builds the API image.
• Now add another service - with the MongoDB image.
• Now try to connect the API to the database, and make the tests to pass.

Need a hint?

Take a look at the solution folder!