dashjson is a tool for exporting (or importing) Datadog dashboards to (or from) json. It allows the Datadog dashboard definitions to be
- version controlled as json files, and
- migrated across Datadog accounts
Create the credentials file containing the api_key and the app_key. Example content of the credentials file (your keys can be found at https://app.datadoghq.com/account/settings#api):
# cat ~/.dashjson.json
{
"api_key": "abcdefg12345678",
"app_key": "abcdefg987654321"
}
Install via pip:
sudo apt install python3-pip
python3 -m pip install --user dashjson
Show usage of the tool:
python3 -m dashjson -h
These are the steps I used on Ubuntu 18.04 with Python 3.6.7 and datadogpy 0.26.0.
Add the following lines to ~/.bashrc:
# Due to https://github.com/pypa/pipenv/issues/2122
export PATH="${HOME}/.local/bin:$PATH"
# Due to https://github.com/pypa/pipenv/issues/1382
export PIPENV_VENV_IN_PROJECT=true
Install pip and pipenv:
sudo apt install python3-pip
python3 -m pip install --user pipenv
Clone and set up the project locally:
git clone git@github.com:ouyi/dashjson.git
cd dashjson
pipenv install --dev
pipenv run python -m unittest
pipenv run python dashjson.py -h
Upload a release:
rm -rf build/ dist/
python setup.py sdist bdist_wheel
pip install . -vvv
python -m dashjson -h
vim setup.py # bump version
git add setup.py && git commit -m "Bump version"
git tag -a my_tag -m "My message"
git push origin master && git push origin --tags
python -m twine upload dist/*
- Automatically determine the dashboard type
For import, the information is available in the input file. For export, we can first get a list of all screenboards and then check existence by id.
- Embed a code coverage badge
The coverage data can be generated with:
coverage run --source=dashjson -m unittest && coverage report && coverage-badge -f -o coverage.svg
- Automate the release process