I'm currently working on a website right now on Django. On my computer, I am running it on Docker with a postgres database. Here's the docker-compose file I have:
version: '3'
services:
db:
image: postgres
environment:
- POSTGRES_DB=postgres
- POSTGRES_USER=postgres
- POSTGRES_PASSWORD=postgres
web:
build: .
volumes:
- .:/usr/src/app
ports:
- "8000:8000"
And here's the relevant part in settings.py
DATABASES = {
'default': {
'ENGINE': 'django.db.backends.postgresql',
'NAME': 'postgres',
'USER': 'postgres',
'PASSWORD': 'postgres',
'HOST': 'db',
'PORT': 5432,
}
}
When I run my tests in the docker container with this setup, it works find and the tests run. However, in github actions, it doesn't work. Here's my workflow file:
name: Django CI
on: push
jobs:
build:
runs-on: ubuntu-latest
strategy:
max-parallel: 4
matrix:
python-version: [3.7, 3.8]
services:
db:
image: postgres
env:
POSTGRES_DB: postgres
POSTGRES_USER: postgres
POSTGRES_PASSWORD: postgres
ports:
- 5432:5432
steps:
- uses: actions/checkout@v2
- name: Set up Python ${{ matrix.python-version }}
uses: actions/setup-python@v1
with:
python-version: ${{ matrix.python-version }}
- name: Install Dependencies
run: |
python -m pip install --upgrade pip
pip install -r requirements.txt
- name: Run Tests
run: |
python manage.py test
When this runs in github actions, I get the following error:
django.db.utils.OperationalError: could not translate host name "db" to address: Temporary failure in name resolution
Could someone please help me with this, and please let me know if you need anymore code.
Django officially supports the following databases: PostgreSQL. MariaDB. MySQL.
PostgreSQL is the default database choice for many Python developers, including the Django team when testing the Django ORM. PostgreSQL is often viewed as more feature robust and stable when compared to MySQL, SQLServer and Oracle. All of those databases are reasonable choices.
adaptor bridges Django's own ORM models to Postgres, but doesn't actually provide the adaptors needed for actual Postgres usage. Rather than re-inventing the wheel, the Django Project has elected to use the psycopg2 library—arguably the most popular python adapter for Postgres.
The reason this is not working is that the settings HOST
is set to db
.
When you're working with a Github Action that only has services, this is called "running all steps on the virtual machine" or VM.
You know this is the case because your steps
are at the same indentation level as services
.
In this situation, you do not specify the service's container name (in this case db
) as the host. Instead, services running in services
as shown in the question are actually running on localhost
which is the same place the steps are taking place.
When the run tests
step is reached, we can see that django is in fact attempting to start up, because the django.db.utils.OperationalError...
error is a Django runtime error. It is showing Django's attempt to reach out and talk to the database.
This means that the settings configuration is wrong. To specifically answer this question, simply update the settings to:
DATABASES = {
'default': {
'ENGINE': 'django.db.backends.postgresql',
'NAME': 'postgres',
'USER': 'postgres',
'PASSWORD': 'postgres',
'HOST': 'localhost',
'PORT': 5432,
}
}
Note a very similar question and resolving answer is shown here.
If this is still not working for you, most likely the settings.py that the Github Actions is running is not the same one you are editing.
As I mention here, ensure first that your setting is not overridden and that your Github Actions job is pulling the correct branch that you are editing the settings.py
.
This can be especially hard because of the delay of waiting for Github Actions to build causes something of a disconnect between what you're doing and the result.
You may think you're working on relevant code / pushing to the right place / setting the correct environment variables when you are not. And you do not detect it possibly because it is easy to get distracted waiting to see if your fix worked. :)
One more note on this, Github has an example simple service using postgres. Look at it carefully. The second example shows the one asked about in this question, where all steps are run on the virtual machine.
As @Rob pointed out the right point, the configuration between settings.py
and PostgreSQL running in the container is wrong. In short, PostgreSQL db in a container does not access your secret key so you need to give them explicitly.
settings.py
DATABASES = {
'default': {
'ENGINE': 'django.db.backends.postgresql',
'NAME': env('POSTGRES_DB'),
'USER': env('POSTGRES_USER'),
'PASSWORD': env('POSTGRES_PASSWORD'),
'HOST': env('POSTGRES_HOST'),
'PORT': 5432
}
}
I use django-environ
to store my environment variables. os.getenv
also works here. Just change os.getenv
with env
.
In your workflow file with the extension of .yml
, services
section should seem like this:
services:
postgres:
image: postgres:12.5
env:
POSTGRES_DB: postgres
POSTGRES_USER: postgres
POSTGRES_PASSWORD: postgres
ports:
- 5432:5432
# needed because the postgres container does not provide a healthcheck
options: --health-cmd pg_isready --health-interval 10s --health-timeout 5s --health-retries 5
In the same file, it is OK to use all environmental variables in the secure way. For example, this shows my Django Testing
workflow:
- name: Django Testing
env:
DEBUG: ${{ secrets.DEBUG }}
SECRET_KEY: ${{ secrets.SECRET_KEY }}
POSTGRES_HOST: ${{ secrets.POSTGRES_HOST }}
POSTGRES_DB: ${{ secrets.POSTGRES_DB }}
POSTGRES_USER: ${{ secrets.POSTGRES_USER }}
POSTGRES_PASSWORD: ${{ secrets.POSTGRES_PASSWORD }}
run: |
python manage.py test
Lastly, in the Github repository secrets, define your variables as :
POSTGRES_HOST --> localhost
POSTGRES_DB --> postgres
POSTGRES_USER --> postgres
POSTGRES_PASSWORD --> postgres
DEBUG --> True
SECRET_KEY --> <Your Django Secret Key>
Finally, you will have:
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With