It’s always better to have a faster tests. For tests which use a database there are two ways to increase the speed:

• a mock library, like mongomock
• in-memory database

With the mock library it’s faster, but it’s more reliable to use real db and usually the mock libraries does not cover all functionality.

For in-memory MongoDB there is an easy way to run it in Docker with a tmpfs volume.

docker run -d --name=mongo_test -p 27018:27017 --tmpfs /data/db mongo:3.4

Then it can be used in tests on 27018 port.

In docker-compose it can be described as:

mongo_test:
    image: mongo:3.4.0
    tmpfs: /data/db

For one of my apps I have an increase more than 3 times, from 84s to 25s.

The same way it also can be used for in-memory MySQL.

docker run --tmpfs /var/lib/mysql -p 3307:3306 -e MYSQL_ROOT_PASSWORD=pass -e MYSQL_DATABASE=db -d --name=mysql-test mysql:5.6

I have an increase also around 3 times, from 3m 33s to 54s.