Testing suite


The tests have been created using Robot Framework and the execution happens inside a Podman container. The only dependency required in the host system is Podman.

Tests structure

All tests are in the tests directory of each NS8 module, in the root of the tests directory the file pythonreq.txt specifies the dependencies required and each sub-directory contains a test suite with the related tests.

The test suite and the related tests are executed in alphabetical order, the __init__.robot can be used for the suite initialization. More details in the Robot Framework documentation.

├── 00_<test_suite>
│   ├── 00_<test>.robot
│   ├── 10_<test>.robot
│   ├── 20_<test>.robot
│   ├── ...
│   ├── ...
│   └── __init__.robot
├── 01_<test_suite>
│   ├── 00_<test>.robot
│   ├── 10_<test>.robot
│   ├── 20_<test>.robot
│   ├── ...
│   ├── ...
│   └── __init__.robot
├── ...
│   ├── ...
│   ├── ...
│   └── __init__.robot
├── pythonreq.txt
└── __init__.robot

Tests execution

Every module has a script named test-module.sh that can be used to setup and launch the tests related to the module. All the tests logs can be found in the directory tests/outputs.


All the scripts have a common way to customize the execution.

./test-module.sh <leader_node> <worker_node1,worker_node2,...>

When lunching the script, make sure the specified SSH_KEYFILE is accessible without password, otherwise the connection to remote host will fail.


  • <leader_node>: The host of the leader node.
  • <worker_node1,worker_node2,...>: A list of comma separated worker hosts.

Environment variables

  • SSH_KEYFILE: SSH private key to use for connection to the NS8 cluster, default ~/.ssh/id_rsa.
  • COREMODULES: list of space-separated module URL to pull and use during the NS8 installation process.

The COREMODULES have the same meaning as the install.sh parameters.

Include/exclude test cases

Test cases can be marked with Robot Framework tags to allow the inclusion and/or exclusion of test cases during tests execution.

Core test suites 10_cluster_sanity and 20_cluster_ui are marked respectively with backend and frontend tags. To execute only backend tests (those included in 10_cluster_sanity) you can edit robot command provided in test-module.sh, adding include parameter:

podman run ... robot ... --include backend ...

The same goes for executing only frontend tests:

podman run ... robot ... --include frontend ...

Some test cases of core module are marked with install or uninstall tag. More precisely:

  • test cases related to NS8 installation and cluster creation are marked with install tag
  • test cases related to NS8 removal are marked with uninstall tag

This pair of tags make it easy to quickly execute tests on a machine with NS8 already installed and configured. Skipping installation, cluster creation and uninstallation drastically reduces tests execution time. To skip NS8 installation, configuration and uninstallation you can edit robot command provided in test-module.sh:

podman run ... robot ... --exclude install --exclude uninstall ...

Headful mode

The execution of UI tests normally happens in headless mode, i.e. without showing the browser window. Anyway, for tests development and debug purposes, seeing test execution in the browser can be very helpful.

At the moment, the only way to do this is to run the tests on your host machine, outside a podman container. Robot Framework installation on your machine is required: the quickest way to install it is using pip pip install -r core/tests/pythonreq.txt, other installations methods are available inside the manual. Then you can run robot command with HEADLESS variable set to false, e.g.:

cd ns8-core
robot -v NODE_ADDR:<LEADER_IP_ADDR> -v HEADLESS:false -v SSH_KEYFILE:$HOME/.ssh/id_rsa -d $HOME/ns8-test-outputs tests/

In the command above we are assuming that:

  • $HOME/.ssh/id_rsa is the SSH identity file to access you NS8 leader machine
  • $HOME/ns8-test-outputs is the location where test report and logs will be written

Testing environment

A terraform configuration for create a clean infrastucture for the tests execution can be found in the ns8-terraform-infra repository.

If you want to use the key generated by Terraform scripts, remember to export it and set SSH_KEYFILE accordingly:

SSH_KEYFILE=../../ns8-terraform-infra/key ./test-module.sh <leader_node> <worker_node1,worker_node2,...>

To access the leader and the nodes with the generated key, use:

ssh -i ../../ns8-terraform-infra/key <leader_node>


Given a domain named test.nethserver.org, running on CentOS Stream 9 on cluster0 workspace, the command to launch the test suites should be:

SSH_KEYFILE=../../ns8-terraform-infra/key ./test-module.sh cs1.leader.cluster0.test.nethserver.org