A docker container which runs the tests from one of the linux system roles against various operating systems.

It starts a virtual machine based on a cloud image of each operating system (defined in images.yml) and runs playbooks in the roles test directory against that machine.

To build the container, run

docker build --tag cockpit/linux-system-roles-test .

To run tests of a role inside the container, execute this inside the top level directory of the test:

docker run --privileged \
           --rm \
           --volume $PWD:/role \
           --volume $CACHEDIR:/cache \
           cockpit/linux-system-roles-test

$CACHEDIR is the directory in which the test downloads virtual machine images. To save some typing, each role contains a script run-tests.sh.

To log into the container, run

docker run -ti --rm --privileged -v $PWD:/role -v $CACHEDIR:/cache \
           cockpit/linux-system-roles-test /bin/bash

You can then run the tests manually with

avocado run test.py -m image:images.yml --show-job-log

To only run tests against a single image, use avocado's parameter filtering mechanism:

avocado run test.py -m image:images.yml --show-job-log \
            --filter-only /run/image/centos-7