Distributed CI environment for development

This repository is used to store the configuration of docker compose in order to provide an environment for development.


To run DCI in docker-compose, first clone this repository:

git clone git@github.com:redhat-cip/dci-dev-env.git

Then, bootstrap dci-dev-env:



Once install, you can launch the environment:

docker-compose -f dci.yml up -d

Then, you can attach containers in order to run parts of the applications:

docker exec -it <container-name> bash

For instance docker exec -it dci-dev-env_api_1 bash


Here is the list of containers for running the application:

  • dci_ansible: contains the dci-ansible code ;
  • dci_api: contains the API of the application and serves it on localhost port 5000, must be started manually ;
  • dci_client: contains the dciclient (Python binding) and dcictl (CLI) clients to DCI Control Server ;
  • dci_db: contains the postgresql database and serves it on localhost port 5432, started by default ;
  • dci_doc: helper to build the project's documentation ;
  • dci_keycloak: keycloak server for SSO. :warning: Use localhost as keycloak's domain as it's set to this value and not ;
  • dci_ui: contains the web app of DCI and serves it on localhost port 8000.
  • dci_swift: storage backend for the API using a Swift server.

API Container

You can initialize or re-initialize the database from the API container by running:

docker exec -it dci-dev-env_api_1 ./bin/dci-dbprovisioning

Client Container

This container allows one to run the python-dciclient within it.

This container is special in several ways compares to the others:

  • It runs systemd ;
  • It runs an sshd daemon that belong to root:root.

Client Usage

To initialize this container you need to install the dciclient library, as well as the agents and feeders:

cd /opt/python-dciclient && pip install --editable ./
cd /opt/python-dciclient/agents && pip install --editable ./
cd /opt/python-dciclient/feeders && pip install --editable ./

Then, Create a local.sh file with the following credentials and source it:

API_CONTAINER_IP="$(docker inspect --format '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}' <CONTAINER_ID>)"
export DCI_LOGIN=admin
export DCI_PASSWORD=admin
export DCI_CS_URL=http://$API_CONTAINER_IP:5000

Keycloak Container

This container allows to use SSO based authentication.


It's provisioned by default by the following:

  • The user admin: username=admin, password=admin ;
  • A realm dci-test ;
  • A client dci-cs ;
  • A lambda user within the dci-test realm: username=dci, password=dci

The client dci-cs is configured to allows OIDC Implicit flow and Direct Access protocols.

Use OIDC Implicit flow

Generate a random value using an UUID, then we will use it for the nonce field in the following link:


This link will redirect the browser to Keycloak SSO login page, once authenticated it will redirect to the redirect_uri value, e.g.:


The id_token parameter correspond to the JSON Web Token (jwt) generated by Keycloak and will be used to authenticated the client on the server.

Use Direct Access protocol

From the CLI:

$ curl \
  --data "client_id=dci-cs" \
  --data "username=dci" \
  --data "password=dci" \
  --data "grant_type=password" \

This will get a JWT and will be used to authenticated the client on the server .

Ansible Container

You can use the Ansible container to run tox:

docker exec -it dci-dev-env_ansible_1 tox

and the functional tests:

docker exec -it dci-dev-env_ansible_1 bash -c 'cd tests; ./run_tests.sh'

Swift Container

If you want to enable the swift storage for the API:

docker-compose -f dci.yml -f dci-swift.yml build
docker-compose -f dci.yml -f dci-swift.yml up -d

Swift is exposed on the non-standard port 5001. If you want to interact with it, you can use your local swift client.

source dci-swift/openrc_dci.sh
swift upload fstab /etc/fstab
swift list

Documentation Container

This container generates DCI documentation.

Use the following command to generate the doc:

docker-compose -f dci.yml -f dci-extra.yml run doc


You will need to install docker-compose and git-review:

pip install -U -r requirements.txt

results matching ""

    No results matching ""