Running docs/vagrant.md example no longer populates the inventory/my_lab/artifacts folder

[oboudry]

Not 100% sure but according to timing this may be a side effect of #8680 also I don't really understand how that fix could have this impact. I raised #8673 for which #8680 was created. Since then the vagrant build as per /docs/Vagrant.md no longer fails, but it is very short and some playbooks are not executed. The inventory/my_lab/artifacts folder is not created, and in the ansible output, the following playbooks are flagged as not running

[WARNING]: Could not match supplied host pattern, ignoring: kube-master

PLAY [Add kube-master nodes to kube_control_plane] *****************************
skipping: no hosts matched
[WARNING]: Could not match supplied host pattern, ignoring: kube-node

PLAY [Add kube-node nodes to kube_node] ****************************************
skipping: no hosts matched
[WARNING]: Could not match supplied host pattern, ignoring: k8s-cluster

PLAY [Add k8s-cluster nodes to k8s_cluster] ************************************
skipping: no hosts matched
[WARNING]: Could not match supplied host pattern, ignoring: calico-rr

PLAY [Add calico-rr nodes to calico_rr] ****************************************
skipping: no hosts matched
[WARNING]: Could not match supplied host pattern, ignoring: no-floating

PLAY [Add no-floating nodes to no_floating] ************************************
skipping: no hosts matched
[WARNING]: Could not match supplied host pattern, ignoring: bastion

PLAY [bastion[0]] **************************************************************
skipping: no hosts matched

Environment:

• Cloud provider or hardware configuration:

• Vagrant running in the exact same conditions as docs/vagrant.md example

• OS (printf "$(uname -srm)\n$(cat /etc/os-release)\n"):
  Linux 5.4.0-107-generic x86_64
  NAME="Ubuntu"
  VERSION="20.04.4 LTS (Focal Fossa)"
  ID=ubuntu
  ID_LIKE=debian
  PRETTY_NAME="Ubuntu 20.04.4 LTS"
  VERSION_ID="20.04"
  HOME_URL="https://www.ubuntu.com/"
  SUPPORT_URL="https://help.ubuntu.com/"
  BUG_REPORT_URL="https://bugs.launchpad.net/ubuntu/"
  PRIVACY_POLICY_URL="https://www.ubuntu.com/legal/terms-and-policies/privacy-policy"
  VERSION_CODENAME=focal
  UBUNTU_CODENAME=focal

• Version of Ansible (ansible --version):
  ansible [core 2.12.3]
  config file = /home/olivier/Documents/kubespray/ansible.cfg
  configured module search path = ['/home/olivier/Documents/kubespray/library']
  ansible python module location = /home/olivier/Documents/kubespray/venv/lib/python3.8/site-packages/ansible
  ansible collection location = /home/olivier/.ansible/collections:/usr/share/ansible/collections
  executable location = /home/olivier/Documents/kubespray/venv/bin/ansible
  python version = 3.8.5 (default, Sep 4 2020, 07:30:14) [GCC 7.3.0]
  jinja version = 2.11.3
  libyaml = True

• Version of Python (python --version):
  Python 3.8.5

Kubespray version (commit) (git rev-parse --short HEAD):
56dda43

Network plugin used:
flannel

Full inventory with variables (ansible -i inventory/sample/inventory.ini all -m debug -a "var=hostvars[inventory_hostname]"):

Command used to invoke ansible:
vagrant provision

Output of ansible run:

https://gist.github.com/oboudry/388b9a22acd2a403589ba48a0b9b5abe

Anything else do we need to know:

[cristicalin]

This was broken more or less on purpose a while back since the vagrantfile was triggering a full deployment and causing our vagrant CI jobs to take too long. Now a vagrant up only calls ansible-playbook --limit all,localhost cluster.yml --tags facts instead of running the full deployment. The changes was #8434 and I'm not quite sure if we should bring it back seeing as the vagrant part is more a support of CI testing and not part of the main functionality of kubespray, this means it has to be fast in CI.

This prevents any artifacts from being created.

[oboudry]

Ok, sorry then, I tried version 2.17.1 which created the artifacts and took it for granted that all version would do.

What I would suggest instead of bring it back is to mention it at the end of the doc and suggest to run the provisionning with full scope, I assume with the following command ?:

ansible-playbook -vvv -i .vagrant/provisioners/ansible/inventory/vagrant_ansible_inventory cluster.yml

[cristicalin]

Yes, that should be the command executed, you may also include --limit all,localhost to emulate how vagrant calls the provisioner.

[oboudry]

Closing, as it's expected behavior.

[oomichi]

I think it would be better to make the deployment succeed if running vagrant up on usual machines as users expect, and good to keep CI just gather facts without actual deployment to make CI stable.