Reproduce CI layout

The CI Framework allows to reproduce the same layout we’re consuming in CI.

While it’s not that different from the install_yamls standard layout (one CRC, one Compute), the CI layout introduces a third node (ansible controller), and all of the nodes have two network interfaces.

The playbooks

You will find reproducer.yml playbook at the root of the project, as well as a specific scenarios/reproducers directory.

A second playbook, reproducer-clean.yml, allows to remove the created resources. Please keep in mind it’s destructive, and you won’t be able to recover anything after the run: virtual machines are destroyed and undefined, and the image layers are removed.

Extra parameter: cifmw_reproducer_repositories

This parameter allows you to pass a list of repositories you want to sync from your local laptop onto the ansible controller machine. The form is pretty easy:

custom/repositories.yml

---
cifmw_reproducer_repositories:
  - src: "{{ playbook_dir }}"
    dest: "src/github.com/openstack-k8s-operators/"
  - src: "{{ cifmw_install_yamls_repo }}"
    dest: "src/github.com/openstack-k8s-operators/"

This one will ensure you have ci_framework as well as install_yamls in a known location on the virtual machine. The dest path matches the one we usually get in CI.

Inventory

This feature can be launched against your own desktop or laptop, but also against a remote hypervisor (preferred due to resources). You can therefore create your own inventory as follows:

custom/inventory.yml

---
all:
  hosts:
    hypervisor:
     ansible_user: your_remote_user
     [any other ansible connection options]

Deploy the layout

Once the layout matches your needs, you just need to run the following:

[laptop]$ ansible-playbook -i custom/inventory.yml \
    reproducer.yml \
    -e cifmw_target_host=hypervisor \
    -e @scenarios/reproducers/networking-definition.yml \
    -e @scenarios/reproducers/3-nodes.yml \
    -e @custom/private-params.yml [-e @custom/repositories.yml]

Deploy succeeds, what’s next?

Once the deploy is over, you will end with ready-to-use virtual machines, provided you passed the needed repositories to sync.

The reproducer injects proper ssh configuration in order to jump on the nodes using their name. Usually, you’ll just need to reach to the “controller”:

[laptop]$ ssh controller-0

You’ll end on the machine, using “zuul” user. Then, you’re all set to run the framework, like we do in the CI.

Cleaning

In order to clean the deployed layout, you can just call the reproducer-clean.yml playbook. It will clean the virtual machines:

[laptop]$ ansible-playbook -i custom/inventory.yml reproducer-clean.yml

It doesn’t require any environment file since it will list all of the existing virtual machines, and clean the ones which name starts with cifmw-. It will also remove the disk images - so if you did set a custom basedir in the deployment, you would need to pass it down accordingly.

Expected layout with the default 3-nodes.yml environment

The provided file should create the following resources on your environment:

[hypervisor]$ virsh net-list --all
 Name            State    Autostart   Persistent
--------------------------------------------------
 cifmw-default   active   no          yes
 cifmw-public    active   no          yes

$ virsh list --all
 Id   Name                 State
------------------------------------
 4    cifmw-compute-0      running
 5    cifmw-controller-0   running
 6    cifmw-crc-0          running

The “Id” value may change.