Merge pull request #12037 from VannTen/ci/convert_vagrant_to_kubevirt_2

CI: convert remaining vagrant jobs (except IPv6) to kubevirt + cleanups

.ansible-lint

@@ -39,5 +39,7 @@ exclude_paths:
   - .github
   - .ansible
   - .cache
+  - .gitlab-ci.yml
+  - .gitlab-ci
 mock_modules:
   - gluster.gluster.gluster_volume

.gitlab-ci.yml

@@ -31,12 +31,12 @@ variables:
   ANSIBLE_VERBOSITY: 2
   RECOVER_CONTROL_PLANE_TEST: "false"
   RECOVER_CONTROL_PLANE_TEST_GROUPS: "etcd[2:]:kube_control_plane[1:]"
-  TERRAFORM_VERSION: 1.3.7
+  TF_VERSION: 1.3.7
   PIPELINE_IMAGE: "$CI_REGISTRY_IMAGE/pipeline:${CI_PIPELINE_ID}-${CI_COMMIT_SHORT_SHA}"
 
 before_script:
   - ./tests/scripts/rebase.sh
-  - mkdir -p /.ssh
+  - mkdir -p cluster-dump $ANSIBLE_INVENTORY
 
 .job: &job
   tags:
@@ -59,18 +59,6 @@ before_script:
     - pre-commit            # lint
     - vagrant-validate      # lint
 
-.testcases: &testcases
-  extends: .job-moderated
-  interruptible: true
-  before_script:
-    - update-alternatives --install /usr/bin/python python /usr/bin/python3 1
-    - ./tests/scripts/rebase.sh
-    - ./tests/scripts/testcases_prepare.sh
-  script:
-    - ./tests/scripts/testcases_run.sh
-  after_script:
-    - ./tests/scripts/testcases_cleanup.sh
-
 # For failfast, at least 1 job must be defined in .gitlab-ci.yml
 # Premoderated with manual actions
 ci-not-authorized:
@@ -102,6 +90,6 @@ include:
   - .gitlab-ci/build.yml
   - .gitlab-ci/lint.yml
   - .gitlab-ci/terraform.yml
-  - .gitlab-ci/packet.yml
+  - .gitlab-ci/kubevirt.yml
   - .gitlab-ci/vagrant.yml
   - .gitlab-ci/molecule.yml

.gitlab-ci/build.yml

@@ -1,5 +1,5 @@
 ---
-.build-container:
+pipeline-image:
   cache:
     key: $CI_COMMIT_REF_SLUG
     paths:
@@ -11,23 +11,19 @@
     name: gcr.io/kaniko-project/executor:debug
     entrypoint: ['']
   variables:
-    TAG: $CI_COMMIT_SHORT_SHA
-    PROJECT_DIR: $CI_PROJECT_DIR
-    DOCKERFILE: Dockerfile
     GODEBUG: "http2client=0"
-  before_script:
+  # TODO: remove the override
-    - echo "{\"auths\":{\"$CI_REGISTRY\":{\"auth\":\"$(echo -n ${CI_REGISTRY_USER}:${CI_REGISTRY_PASSWORD} | base64)\"}}}" > /kaniko/.docker/config.json
+  # currently rebase.sh depends on bash (not available in the kaniko image)
+  # once we have a simpler rebase (which should be easy if the target branch ref is available as variable
+  # we'll be able to rebase here as well hopefully
+  before_script: []
   script:
+    - echo "{\"auths\":{\"$CI_REGISTRY\":{\"auth\":\"$(echo -n ${CI_REGISTRY_USER}:${CI_REGISTRY_PASSWORD} | base64)\"}}}" > /kaniko/.docker/config.json
     - /kaniko/executor --cache=true
                        --cache-dir=image-cache
-                       --context $PROJECT_DIR
+                       --context $CI_PROJECT_DIR
-                       --dockerfile $PROJECT_DIR/$DOCKERFILE
+                       --dockerfile $CI_PROJECT_DIR/pipeline.Dockerfile
                        --label 'git-branch'=$CI_COMMIT_REF_SLUG
                        --label 'git-tag=$CI_COMMIT_TAG'
                        --destination $PIPELINE_IMAGE
                        --log-timestamp=true
-
-pipeline-image:
-  extends: .build-container
-  variables:
-    DOCKERFILE: pipeline.Dockerfile

.gitlab-ci/kubevirt.yml

@@ -0,0 +1,147 @@
+---
+.kubevirt:
+  extends: .job-moderated
+  interruptible: true
+  script:
+    - ansible-playbook tests/cloud_playbooks/create-kubevirt.yml
+                      -c local -e @"tests/files/${TESTCASE}.yml"
+    - ./tests/scripts/testcases_run.sh
+  variables:
+    ANSIBLE_TIMEOUT: "120"
+  tags:
+    - ffci
+  needs:
+    - pipeline-image
+    - ci-not-authorized
+
+# TODO: generate testcases matrixes from the files in tests/files/
+# this is needed to avoid the need for PR rebasing when a job was added or remvoed in the target branch
+# (currently, a removed job in the target branch breaks the tests, because the
+# pipeline definition is parsed by gitlab before the rebase.sh script)
+# CI template for PRs
+pr:
+  stage: deploy-part1
+  rules:
+    - if: $PR_LABELS =~ /.*ci-short.*/
+      when: manual
+      allow_failure: true
+    - if: $CI_COMMIT_BRANCH =~ /^pr-.*$/
+      when: on_success
+    - when: manual
+      allow_failure: true
+  extends: .kubevirt
+  parallel:
+    matrix:
+      - TESTCASE:
+          - almalinux8-calico
+          - almalinux9-crio
+          - almalinux9-kube-ovn
+          - debian11-calico-collection
+          - debian11-macvlan
+          - debian12-cilium
+          - fedora39-kube-router
+            # FIXME: this test if broken (perma-failing)
+          - openeuler24-calico
+          - opensuse15-6-calico
+          - rockylinux8-calico
+          - rockylinux9-cilium
+          - ubuntu20-calico-all-in-one-hardening
+          - ubuntu20-cilium-sep
+          - ubuntu20-flannel-collection
+          - ubuntu20-kube-router-sep
+          - ubuntu20-kube-router-svc-proxy
+          - ubuntu22-calico-all-in-one
+          - ubuntu22-calico-all-in-one-upgrade
+          - ubuntu24-calico-etcd-datastore
+
+# The ubuntu20-calico-all-in-one jobs are meant as early stages to prevent running the full CI if something is horribly broken
+ubuntu20-calico-all-in-one:
+  stage: deploy-part1
+  extends: .kubevirt
+  variables:
+    TESTCASE: ubuntu20-calico-all-in-one
+  rules:
+    - if: $CI_COMMIT_BRANCH =~ /^pr-.*$/
+      when: on_success
+    - when: manual
+      allow_failure: true
+
+pr_full:
+  extends: .kubevirt
+  stage: deploy-extended
+  rules:
+    - if: $PR_LABELS =~ /.*ci-full.*/
+      when: on_success
+    # Else run as manual
+    - when: manual
+      allow_failure: true
+  parallel:
+    matrix:
+      - TESTCASE:
+          - almalinux9-calico-ha-ebpf
+          - almalinux9-calico-nodelocaldns-secondary
+          - debian11-custom-cni
+          - debian11-kubelet-csr-approver
+          - debian12-custom-cni-helm
+          - fedora39-calico-swap-selinux
+          - fedora39-crio
+          - ubuntu20-all-in-one-docker
+          - ubuntu20-calico-ha-wireguard
+          - ubuntu20-flannel-ha
+          - ubuntu20-flannel-ha-once
+
+# Need an update of the container image to use schema v2
+# update: quay.io/kubespray/vm-amazon-linux-2:latest
+manual:
+  extends: pr_full
+  parallel:
+    matrix:
+      - TESTCASE:
+          - amazon-linux-2-all-in-one
+  rules:
+    - when: manual
+      allow_failure: true
+
+pr_extended:
+  extends: .kubevirt
+  stage: deploy-extended
+  rules:
+    - if: $PR_LABELS =~ /.*(ci-extended|ci-full).*/
+      when: on_success
+    - when: manual
+      allow_failure: true
+  parallel:
+    matrix:
+      - TESTCASE:
+          - almalinux9-calico
+          - almalinux9-calico-remove-node
+          - almalinux9-docker
+          - debian11-docker
+          - debian12-calico
+          - debian12-docker
+          - opensuse15-6-docker-cilium
+          - rockylinux9-calico
+          - ubuntu20-calico-etcd-kubeadm
+          - ubuntu20-flannel
+          - ubuntu22-all-in-one-docker
+          - ubuntu24-all-in-one-docker
+          - ubuntu24-calico-all-in-one
+
+# Enabled when PERIODIC_CI_ENABLED var is set
+periodic:
+  only:
+    variables:
+      - $PERIODIC_CI_ENABLED
+  allow_failure: true
+  extends: .kubevirt
+  parallel:
+    matrix:
+      - TESTCASE:
+          - debian11-calico-upgrade
+          - debian11-calico-upgrade-once
+          - debian12-cilium-svc-proxy
+          - fedora39-calico-selinux
+          - fedora40-docker-calico
+          - ubuntu20-calico-etcd-kubeadm-upgrade-ha
+          - ubuntu20-calico-ha-recover
+          - ubuntu20-calico-ha-recover-noquorum

.gitlab-ci/molecule.yml

@@ -8,8 +8,6 @@
   needs:
   - pipeline-image
   # - ci-not-authorized
-  before_script:
-  - ./tests/scripts/rebase.sh
   script:
   - ./tests/scripts/molecule_run.sh
   after_script:

.gitlab-ci/packet.yml

@@ -1,257 +0,0 @@
----
-.packet:
-  extends: .testcases
-  variables:
-    ANSIBLE_TIMEOUT: "120"
-    CI_PLATFORM: packet
-    SSH_USER: kubespray
-  tags:
-    - ffci
-  needs:
-    - pipeline-image
-    - ci-not-authorized
-
-# CI template for PRs
-.packet_pr:
-  stage: deploy-part1
-  rules:
-    - if: $PR_LABELS =~ /.*ci-short.*/
-      when: manual
-      allow_failure: true
-    - if: $CI_COMMIT_BRANCH =~ /^pr-.*$/
-      when: on_success
-    - when: manual
-      allow_failure: true
-  extends: .packet
-
-  ## Uncomment this to have multiple stages
-  # needs:
-  #   - packet_ubuntu20-calico-all-in-one
-
-.packet_pr_short:
-  stage: deploy-part1
-  extends: .packet
-  rules:
-    - if: $CI_COMMIT_BRANCH =~ /^pr-.*$/
-      when: on_success
-    - when: manual
-      allow_failure: true
-
-.packet_pr_manual:
-  extends: .packet_pr
-  stage: deploy-extended
-  rules:
-    - if: $PR_LABELS =~ /.*ci-full.*/
-      when: on_success
-    # Else run as manual
-    - when: manual
-      allow_failure: true
-
-.packet_pr_extended:
-  extends: .packet_pr
-  stage: deploy-extended
-  rules:
-    - if: $PR_LABELS =~ /.*(ci-extended|ci-full).*/
-      when: on_success
-    - when: manual
-      allow_failure: true
-
-# CI template for periodic CI jobs
-# Enabled when PERIODIC_CI_ENABLED var is set
-.packet_periodic:
-  only:
-    variables:
-      - $PERIODIC_CI_ENABLED
-  allow_failure: true
-  extends: .packet
-
-# The ubuntu20-calico-all-in-one jobs are meant as early stages to prevent running the full CI if something is horribly broken
-packet_ubuntu20-calico-all-in-one:
-  stage: deploy-part1
-  extends: .packet_pr_short
-  variables:
-    RESET_CHECK: "true"
-
-# ### PR JOBS PART2
-
-packet_ubuntu20-crio:
-  extends: .packet_pr_manual
-
-packet_ubuntu22-calico-all-in-one:
-  extends: .packet_pr
-
-packet_ubuntu22-calico-all-in-one-upgrade:
-  extends: .packet_pr
-  variables:
-    UPGRADE_TEST: graceful
-
-packet_ubuntu24-calico-etcd-datastore:
-  extends: .packet_pr
-
-packet_almalinux9-crio:
-  extends: .packet_pr
-
-packet_almalinux9-kube-ovn:
-  extends: .packet_pr
-
-packet_debian11-calico-collection:
-  extends: .packet_pr
-
-packet_debian11-macvlan:
-  extends: .packet_pr
-
-packet_debian12-cilium:
-  extends: .packet_pr
-
-packet_almalinux8-calico:
-  extends: .packet_pr
-
-packet_rockylinux8-calico:
-  extends: .packet_pr
-
-packet_rockylinux9-cilium:
-  extends: .packet_pr
-  variables:
-    RESET_CHECK: "true"
-
-# Need an update of the container image to use schema v2
-# update: quay.io/kubespray/vm-amazon-linux-2:latest
-packet_amazon-linux-2-all-in-one:
-  extends: .packet_pr_manual
-  rules:
-    - when: manual
-      allow_failure: true
-
-packet_opensuse15-6-calico:
-  extends: .packet_pr
-
-packet_ubuntu20-cilium-sep:
-  extends: .packet_pr
-
-packet_openeuler24-calico:
-  extends: .packet_pr
-
-packet_ubuntu20-calico-all-in-one-hardening:
-  extends: .packet_pr
-
-## Extended
-packet_debian11-docker:
-  extends: .packet_pr_extended
-
-packet_debian12-docker:
-  extends: .packet_pr_extended
-
-packet_debian12-calico:
-  extends: .packet_pr_extended
-
-packet_almalinux9-calico-remove-node:
-  extends: .packet_pr_extended
-  variables:
-    REMOVE_NODE_CHECK: "true"
-    REMOVE_NODE_NAME: "instance-3"
-
-packet_rockylinux9-calico:
-  extends: .packet_pr_extended
-
-packet_almalinux9-calico:
-  extends: .packet_pr_extended
-
-packet_almalinux9-docker:
-  extends: .packet_pr_extended
-
-packet_opensuse15-6-docker-cilium:
-  extends: .packet_pr_extended
-
-packet_ubuntu24-calico-all-in-one:
-  extends: .packet_pr_extended
-
-packet_ubuntu20-calico-etcd-kubeadm:
-  extends: .packet_pr_extended
-
-packet_ubuntu24-all-in-one-docker:
-  extends: .packet_pr_extended
-
-packet_ubuntu22-all-in-one-docker:
-  extends: .packet_pr_extended
-
-# ### MANUAL JOBS
-packet_fedora39-crio:
-  extends: .packet_pr_manual
-
-packet_ubuntu20-flannel-ha:
-  extends: .packet_pr_manual
-
-packet_ubuntu20-all-in-one-docker:
-  extends: .packet_pr_manual
-
-packet_ubuntu20-flannel-ha-once:
-  extends: .packet_pr_manual
-
-packet_fedora39-calico-swap-selinux:
-  extends: .packet_pr_manual
-
-packet_almalinux9-calico-ha-ebpf:
-  extends: .packet_pr_manual
-
-packet_almalinux9-calico-nodelocaldns-secondary:
-  extends: .packet_pr_manual
-
-packet_debian11-custom-cni:
-  extends: .packet_pr_manual
-
-packet_debian11-kubelet-csr-approver:
-  extends: .packet_pr_manual
-
-packet_debian12-custom-cni-helm:
-  extends: .packet_pr_manual
-
-packet_ubuntu20-calico-ha-wireguard:
-  extends: .packet_pr_manual
-
-# PERIODIC
-packet_fedora40-docker-calico:
-  stage: deploy-extended
-  extends: .packet_periodic
-  variables:
-    RESET_CHECK: "true"
-
-packet_fedora39-calico-selinux:
-  stage: deploy-extended
-  extends: .packet_periodic
-
-packet_ubuntu20-calico-etcd-kubeadm-upgrade-ha:
-  stage: deploy-extended
-  extends: .packet_periodic
-  variables:
-    UPGRADE_TEST: basic
-
-
-packet_debian11-calico-upgrade-once:
-  stage: deploy-extended
-  extends: .packet_periodic
-  variables:
-    UPGRADE_TEST: graceful
-
-packet_ubuntu20-calico-ha-recover:
-  stage: deploy-extended
-  extends: .packet_periodic
-  variables:
-    RECOVER_CONTROL_PLANE_TEST: "true"
-    RECOVER_CONTROL_PLANE_TEST_GROUPS: "etcd[2:]:kube_control_plane[1:]"
-
-packet_ubuntu20-calico-ha-recover-noquorum:
-  stage: deploy-extended
-  extends: .packet_periodic
-  variables:
-    RECOVER_CONTROL_PLANE_TEST: "true"
-    RECOVER_CONTROL_PLANE_TEST_GROUPS: "etcd[1:]:kube_control_plane[1:]"
-
-packet_debian11-calico-upgrade:
-  stage: deploy-extended
-  extends: .packet_periodic
-  variables:
-    UPGRADE_TEST: graceful
-
-packet_debian12-cilium-svc-proxy:
-  stage: deploy-extended
-  extends: .packet_periodic

.gitlab-ci/terraform.yml

@@ -5,28 +5,21 @@
   needs:
     - ci-not-authorized
     - pipeline-image
+  variables:
+    TF_VAR_public_key_path: "${ANSIBLE_PRIVATE_KEY_FILE}.pub"
+    TF_VAR_ssh_private_key_path: $ANSIBLE_PRIVATE_KEY_FILE
+    CLUSTER: $CI_COMMIT_REF_NAME
+    TERRAFORM_STATE_ROOT: $CI_PROJECT_DIR
   stage: deploy-part1
   before_script:
-    - update-alternatives --install /usr/bin/python python /usr/bin/python3 1
     - ./tests/scripts/rebase.sh
-    - ./tests/scripts/testcases_prepare.sh
+    - mkdir -p cluster-dump $ANSIBLE_INVENTORY
     - ./tests/scripts/terraform_install.sh
-    # Set Ansible config
-    - cp ansible.cfg ~/.ansible.cfg
-    # Prepare inventory
     - cp contrib/terraform/$PROVIDER/sample-inventory/cluster.tfvars .
-    - ln -s contrib/terraform/$PROVIDER/hosts
+    - ln -rs -t $ANSIBLE_INVENTORY contrib/terraform/$PROVIDER/hosts
     - terraform -chdir="contrib/terraform/$PROVIDER" init
-    # Copy SSH keypair
-    - mkdir -p ~/.ssh
-    - echo "$PACKET_PRIVATE_KEY" | base64 -d > ~/.ssh/id_rsa
-    - chmod 400 ~/.ssh/id_rsa
-    - echo "$PACKET_PUBLIC_KEY" | base64 -d > ~/.ssh/id_rsa.pub
-    - mkdir -p contrib/terraform/$PROVIDER/group_vars
-    # Random subnet to avoid routing conflicts
-    - export TF_VAR_subnet_cidr="10.$(( $RANDOM % 256 )).$(( $RANDOM % 256 )).0/24"
 
-.terraform_validate:
+terraform_validate:
   extends: .terraform_install
   tags: [ffci]
   only: ['master', /^pr-.*$/]
@@ -36,6 +29,17 @@
   stage: test
   needs:
     - pipeline-image
+  parallel:
+    matrix:
+      - PROVIDER:
+          - openstack
+          - equinix
+          - aws
+          - exoscale
+          - hetzner
+          - vsphere
+          - upcloud
+          - nifcloud
 
 .terraform_apply:
   extends: .terraform_install
@@ -43,99 +47,24 @@
   stage: deploy-extended
   when: manual
   only: [/^pr-.*$/]
-  artifacts:
-    when: always
-    paths:
-      - cluster-dump/
   variables:
     ANSIBLE_INVENTORY_UNPARSED_FAILED: "true"
-    ANSIBLE_INVENTORY: hosts
+    ANSIBLE_REMOTE_USER: ubuntu # the openstack terraform module does not handle custom user correctly
-    CI_PLATFORM: tf
+    ANSIBLE_SSH_RETRIES: 15
-    TF_VAR_ssh_user: $SSH_USER
+    TF_VAR_ssh_user: $ANSIBLE_REMOTE_USER
     TF_VAR_cluster_name: $CI_JOB_ID
   script:
+    # Set Ansible config
+    - cp ansible.cfg ~/.ansible.cfg
+    - ssh-keygen -N '' -f $ANSIBLE_PRIVATE_KEY_FILE -t rsa
+    - mkdir -p contrib/terraform/$PROVIDER/group_vars
+    # Random subnet to avoid routing conflicts
+    - export TF_VAR_subnet_cidr="10.$(( $RANDOM % 256 )).$(( $RANDOM % 256 )).0/24"
+    - terraform -chdir="contrib/terraform/$PROVIDER" apply -auto-approve -parallelism=1
     - tests/scripts/testcases_run.sh
   after_script:
     # Cleanup regardless of exit code
-    - ./tests/scripts/testcases_cleanup.sh
+    - terraform -chdir="contrib/terraform/$PROVIDER" destroy -auto-approve
-
-tf-validate-openstack:
-  extends: .terraform_validate
-  variables:
-    TF_VERSION: $TERRAFORM_VERSION
-    PROVIDER: openstack
-    CLUSTER: $CI_COMMIT_REF_NAME
-
-tf-validate-equinix:
-  extends: .terraform_validate
-  variables:
-    TF_VERSION: $TERRAFORM_VERSION
-    PROVIDER: equinix
-    CLUSTER: $CI_COMMIT_REF_NAME
-
-tf-validate-aws:
-  extends: .terraform_validate
-  variables:
-    TF_VERSION: $TERRAFORM_VERSION
-    PROVIDER: aws
-    CLUSTER: $CI_COMMIT_REF_NAME
-
-tf-validate-exoscale:
-  extends: .terraform_validate
-  variables:
-    TF_VERSION: $TERRAFORM_VERSION
-    PROVIDER: exoscale
-
-tf-validate-hetzner:
-  extends: .terraform_validate
-  variables:
-    TF_VERSION: $TERRAFORM_VERSION
-    PROVIDER: hetzner
-
-tf-validate-vsphere:
-  extends: .terraform_validate
-  variables:
-    TF_VERSION: $TERRAFORM_VERSION
-    PROVIDER: vsphere
-    CLUSTER: $CI_COMMIT_REF_NAME
-
-tf-validate-upcloud:
-  extends: .terraform_validate
-  variables:
-    TF_VERSION: $TERRAFORM_VERSION
-    PROVIDER: upcloud
-    CLUSTER: $CI_COMMIT_REF_NAME
-
-tf-validate-nifcloud:
-  extends: .terraform_validate
-  variables:
-    TF_VERSION: $TERRAFORM_VERSION
-    PROVIDER: nifcloud
-
-# tf-packet-ubuntu20-default:
-#   extends: .terraform_apply
-#   variables:
-#     TF_VERSION: $TERRAFORM_VERSION
-#     PROVIDER: packet
-#     CLUSTER: $CI_COMMIT_REF_NAME
-#     TF_VAR_number_of_k8s_masters: "1"
-#     TF_VAR_number_of_k8s_nodes: "1"
-#     TF_VAR_plan_k8s_masters: t1.small.x86
-#     TF_VAR_plan_k8s_nodes: t1.small.x86
-#     TF_VAR_metro: am
-#     TF_VAR_public_key_path: ""
-#     TF_VAR_operating_system: ubuntu_20_04
-
-.ovh_variables: &ovh_variables
-  OS_AUTH_URL: https://auth.cloud.ovh.net/v3
-  OS_PROJECT_ID: 8d3cd5d737d74227ace462dee0b903fe
-  OS_PROJECT_NAME: "9361447987648822"
-  OS_USER_DOMAIN_NAME: Default
-  OS_PROJECT_DOMAIN_ID: default
-  OS_USERNAME: 8XuhBMfkKVrk
-  OS_REGION_NAME: UK1
-  OS_INTERFACE: public
-  OS_IDENTITY_API_VERSION: "3"
 
 # Elastx is generously donating resources for Kubespray on Openstack CI
 # Contacts: @gix @bl0m1
@@ -169,11 +98,8 @@ tf-elastx_ubuntu20-calico:
   allow_failure: true
   variables:
     <<: *elastx_variables
-    TF_VERSION: $TERRAFORM_VERSION
     PROVIDER: openstack
-    CLUSTER: $CI_COMMIT_REF_NAME
     ANSIBLE_TIMEOUT: "60"
-    SSH_USER: ubuntu
     TF_VAR_number_of_k8s_masters: "1"
     TF_VAR_number_of_k8s_masters_no_floating_ip: "0"
     TF_VAR_number_of_k8s_masters_no_floating_ip_no_etcd: "0"
@@ -194,46 +120,3 @@ tf-elastx_ubuntu20-calico:
     TF_VAR_flavor_k8s_node: 3f73fc93-ec61-4808-88df-2580d94c1a9b      # v1-standard-2
     TF_VAR_image: ubuntu-20.04-server-latest
     TF_VAR_k8s_allowed_remote_ips: '["0.0.0.0/0"]'
-
-# OVH voucher expired, commenting job until things are sorted  out
-
-# tf-ovh_cleanup:
-#  stage: unit-tests
-#  tags: [light]
-#  image: python
-#  environment: ovh
-#  variables:
-#    <<: *ovh_variables
-#  before_script:
-#    - pip install -r scripts/openstack-cleanup/requirements.txt
-#  script:
-#    - ./scripts/openstack-cleanup/main.py
-
-# tf-ovh_ubuntu20-calico:
-#  extends: .terraform_apply
-#  when: on_success
-#  environment: ovh
-#  variables:
-#    <<: *ovh_variables
-#    TF_VERSION: $TERRAFORM_VERSION
-#    PROVIDER: openstack
-#    CLUSTER: $CI_COMMIT_REF_NAME
-#    ANSIBLE_TIMEOUT: "60"
-#    SSH_USER: ubuntu
-#    TF_VAR_number_of_k8s_masters: "0"
-#    TF_VAR_number_of_k8s_masters_no_floating_ip: "1"
-#    TF_VAR_number_of_k8s_masters_no_floating_ip_no_etcd: "0"
-#    TF_VAR_number_of_etcd: "0"
-#    TF_VAR_number_of_k8s_nodes: "0"
-#    TF_VAR_number_of_k8s_nodes_no_floating_ip: "1"
-#    TF_VAR_number_of_gfs_nodes_no_floating_ip: "0"
-#    TF_VAR_number_of_bastions: "0"
-#    TF_VAR_number_of_k8s_masters_no_etcd: "0"
-#    TF_VAR_use_neutron: "0"
-#    TF_VAR_floatingip_pool: "Ext-Net"
-#    TF_VAR_external_net: "6011fbc9-4cbf-46a4-8452-6890a340b60b"
-#    TF_VAR_network_name: "Ext-Net"
-#    TF_VAR_flavor_k8s_master: "defa64c3-bd46-43b4-858a-d93bbae0a229"    # s1-8
-#    TF_VAR_flavor_k8s_node: "defa64c3-bd46-43b4-858a-d93bbae0a229"      # s1-8
-#    TF_VAR_image: "Ubuntu 20.04"
-#    TF_VAR_k8s_allowed_remote_ips: '["0.0.0.0/0"]'

.gitlab-ci/vagrant.yml

@@ -1,13 +1,13 @@
 ---
-.vagrant:
+vagrant:
-  extends: .testcases
+  extends: .job-moderated
   needs:
     - ci-not-authorized
   variables:
     CI_PLATFORM: "vagrant"
     SSH_USER: "vagrant"
     VAGRANT_DEFAULT_PROVIDER: "libvirt"
-    KUBESPRAY_VAGRANT_CONFIG: tests/files/${CI_JOB_NAME}.rb
+    KUBESPRAY_VAGRANT_CONFIG: tests/files/${TESTCASE}.rb
     DOCKER_NAME: vagrant
     VAGRANT_ANSIBLE_TAGS: facts
     VAGRANT_HOME: "$CI_PROJECT_DIR/.vagrant.d"
@@ -28,54 +28,22 @@
     - pip install --no-compile --no-cache-dir -r $CI_PROJECT_DIR/tests/requirements.txt
     - ./tests/scripts/vagrant_clean.sh
   script:
+    - vagrant up
     - ./tests/scripts/testcases_run.sh
+  after_script:
+    - vagrant destroy -f
   cache:
     key: $CI_JOB_NAME_SLUG
     paths:
       - .vagrant.d/boxes
       - .cache/pip
     policy: pull-push # TODO: change to "pull" when not on main
-
-vagrant_ubuntu24-calico-dual-stack:
   stage: deploy-extended
-  extends: .vagrant
   rules:
     - if: $PR_LABELS =~ /.*(ci-extended|ci-full).*/
       when: on_success
-  allow_failure: false
+  parallel:
-
+    matrix:
-vagrant_ubuntu24-calico-ipv6only-stack:
+      - TESTCASE:
-  stage: deploy-extended
+          - ubuntu24-calico-dual-stack
-  extends: .vagrant
+          - ubuntu24-calico-ipv6only-stack
-  rules:
-    - if: $PR_LABELS =~ /.*(ci-extended|ci-full).*/
-      when: on_success
-  allow_failure: false
-
-vagrant_ubuntu20-flannel:
-  stage: deploy-part1
-  extends: .vagrant
-  when: on_success
-  allow_failure: false
-
-vagrant_ubuntu20-flannel-collection:
-  stage: deploy-extended
-  extends: .vagrant
-  when: manual
-
-vagrant_ubuntu20-kube-router-sep:
-  stage: deploy-extended
-  extends: .vagrant
-  when: manual
-
-# Service proxy test fails connectivity testing
-vagrant_ubuntu20-kube-router-svc-proxy:
-  stage: deploy-extended
-  extends: .vagrant
-  when: manual
-
-vagrant_fedora39-kube-router:
-  stage: deploy-extended
-  extends: .vagrant
-  when: manual
-# FIXME: this test if broken (perma-failing)

roles/adduser/molecule/default/molecule.yml

@@ -14,6 +14,6 @@ provisioner:
       callbacks_enabled: profile_tasks
       timeout: 120
   playbooks:
-    create: ../../../../tests/cloud_playbooks/create-packet.yml
+    create: ../../../../tests/cloud_playbooks/create-kubevirt.yml
 verifier:
   name: testinfra

roles/bastion-ssh-config/molecule/default/molecule.yml

@@ -22,6 +22,6 @@ provisioner:
             hosts:
               bastion-01:
   playbooks:
-    create: ../../../../tests/cloud_playbooks/create-packet.yml
+    create: ../../../../tests/cloud_playbooks/create-kubevirt.yml
 verifier:
   name: testinfra

roles/bootstrap-os/molecule/default/molecule.yml

@@ -32,6 +32,6 @@ provisioner:
           name: foo
           comment: My test comment
   playbooks:
-    create: ../../../../tests/cloud_playbooks/create-packet.yml
+    create: ../../../../tests/cloud_playbooks/create-kubevirt.yml
 verifier:
   name: testinfra

roles/container-engine/containerd/molecule/default/molecule.yml

@@ -34,6 +34,6 @@ provisioner:
       callbacks_enabled: profile_tasks
       timeout: 120
   playbooks:
-    create: ../../../../../tests/cloud_playbooks/create-packet.yml
+    create: ../../../../../tests/cloud_playbooks/create-kubevirt.yml
 verifier:
   name: testinfra

roles/container-engine/cri-dockerd/molecule/default/molecule.yml

@@ -26,6 +26,6 @@ provisioner:
       all:
         become: true
   playbooks:
-    create: ../../../../../tests/cloud_playbooks/create-packet.yml
+    create: ../../../../../tests/cloud_playbooks/create-kubevirt.yml
 verifier:
   name: testinfra

roles/container-engine/cri-o/molecule/default/molecule.yml

@@ -42,6 +42,6 @@ provisioner:
       callbacks_enabled: profile_tasks
       timeout: 120
   playbooks:
-    create: ../../../../../tests/cloud_playbooks/create-packet.yml
+    create: ../../../../../tests/cloud_playbooks/create-kubevirt.yml
 verifier:
   name: testinfra

tests/Makefile

@@ -8,7 +8,7 @@ $(ANSIBLE_INVENTORY):
 	mkdir $@
 
 create-packet: | $(ANSIBLE_INVENTORY)
-	ansible-playbook cloud_playbooks/create-packet.yml -c local \
+	ansible-playbook cloud_playbooks/create-kubevirt.yml -c local \
 		-e @"files/${CI_JOB_NAME}.yml"
 
 delete-packet: ;

tests/cloud_playbooks/roles/packet-ci/tasks/main.yml

@@ -1,8 +1,4 @@
 ---
-- name: Include custom vars for ci job
-  include_vars: "../files/{{ lookup('ansible.builtin.env', 'CI_JOB_NAME') }}.yml"
-  when: molecule_yml is not defined
-
 - name: Generate SSH keypair
   community.crypto.openssh_keypair:
     size: 2048

tests/files/almalinux9-calico-remove-node

@@ -0,0 +1,2 @@
+REMOVE_NODE_CHECK=true
+REMOVE_NODE_NAME=instance-3

tests/files/debian11-calico-upgrade

@@ -0,0 +1 @@
+UPGRADE_TEST=graceful

tests/files/debian11-calico-upgrade-once

@@ -0,0 +1 @@
+UPGRADE_TEST=graceful

tests/files/fedora39-kube-router.yml

@@ -0,0 +1,7 @@
+---
+cloud_image: fedora-39
+cluster_layout:
+  - node_groups: ['kube_control_plane', 'etcd', 'kube_node']
+  - node_groups: ['kube_node']
+
+kube_network_plugin: "kube-router"

tests/files/fedora40-docker.calico

@@ -0,0 +1 @@
+RESET_CHECK=true

tests/files/rockylinux9-cilium

@@ -0,0 +1 @@
+RESET_CHECK=true

tests/files/ubuntu20-calico-all-in-one

@@ -0,0 +1 @@
+RESET_CHECK=true

tests/files/ubuntu20-calico-etcd-kubeadm-upgrade-ha

@@ -0,0 +1 @@
+UPGRADE_TEST=basic

tests/files/ubuntu20-calico-ha-recover

@@ -0,0 +1,2 @@
+RECOVER_CONTROL_PLANE_TEST=true
+RECOVER_CONTROL_PLANE_TEST_GROUPS="etcd[2:]:kube_control_plane[1:]"

tests/files/ubuntu20-calico-ha-recover-noquorum

@@ -0,0 +1,2 @@
+RECOVER_CONTROL_PLANE_TEST=true
+RECOVER_CONTROL_PLANE_TEST_GROUPS="etcd[1:]:kube_control_plane[1:]"

tests/files/ubuntu20-flannel-collection.yml

@@ -0,0 +1,8 @@
+---
+cloud_image: ubuntu-2004
+cluster_layout:
+  - node_groups: ['kube_control_plane', 'etcd', 'kube_node']
+  - node_groups: ['kube_control_plane', 'etcd', 'kube_node']
+  - node_groups: ['etcd', 'kube_node']
+
+kube_network_plugin: flannel

tests/files/ubuntu20-flannel.yml

@@ -0,0 +1 @@
+ubuntu20-flannel-collection.yml

tests/files/ubuntu20-kube-router-sep.yml

@@ -0,0 +1,7 @@
+---
+cloud_image: ubuntu-2004
+cluster_layout:
+  - node_groups: ['kube_control_plane', 'etcd', 'kube_node']
+  - node_groups: ['kube_node']
+
+kube_network_plugin: "kube-router"

tests/files/ubuntu20-kube-router-svc-proxy.yml

@@ -0,0 +1,10 @@
+---
+cloud_image: ubuntu-2004
+cluster_layout:
+  - node_groups: ['kube_control_plane', 'etcd', 'kube_node']
+  - node_groups: ['kube_control_plane', 'etcd', 'kube_node']
+  - node_groups: ['etcd', 'kube_node']
+
+kube_network_plugin: "kube-router"
+
+kube_router_run_service_proxy: true

tests/files/ubuntu22-calico-all-in-one-upgrade

@@ -0,0 +1 @@
+UPGRADE_TEST=graceful

tests/files/vagrant_fedora39-kube-router.rb

@@ -1,15 +0,0 @@
-$num_instances = 2
-$vm_memory ||= 2048
-$os = "fedora39"
-
-$control_plane_instances = 1
-$etcd_instances = 1
-
-# For CI we are not worried about data persistence across reboot
-$libvirt_volume_cache = "unsafe"
-
-# Checking for box update can trigger API rate limiting
-# https://www.vagrantup.com/docs/vagrant-cloud/request-limits.html
-$box_check_update = false
-
-$network_plugin = "kube-router"

tests/files/vagrant_fedora39-kube-router.yml

@@ -1,6 +0,0 @@
----
-# Instance settings
-cloud_image: fedora-39
-
-# Kubespray settings
-kube_network_plugin: kube-router

tests/files/vagrant_ubuntu20-flannel-collection.rb

@@ -1,9 +0,0 @@
-$os = "ubuntu2004"
-
-# For CI we are not worries about data persistence across reboot
-$libvirt_volume_cache = "unsafe"
-
-# Checking for box update can trigger API rate limiting
-# https://www.vagrantup.com/docs/vagrant-cloud/request-limits.html
-$box_check_update = false
-$vm_cpus = 2

tests/files/vagrant_ubuntu20-flannel-collection.yml

@@ -1,3 +0,0 @@
----
-# Kubespray settings
-kube_network_plugin: flannel

tests/files/vagrant_ubuntu20-flannel.rb

@@ -1,8 +0,0 @@
-$os = "ubuntu2004"
-
-# For CI we are not worries about data persistence across reboot
-$libvirt_volume_cache = "unsafe"
-# Checking for box update can trigger API rate limiting
-# https://www.vagrantup.com/docs/vagrant-cloud/request-limits.html
-$box_check_update = false
-$vm_cpus = 2

tests/files/vagrant_ubuntu20-flannel.yml

@@ -1,4 +0,0 @@
----
-# Kubespray settings
-kube_network_plugin: flannel
-ansible_ssh_private_key: .vagrant.d/insecure_private_key

tests/files/vagrant_ubuntu20-kube-router-sep.rb

@@ -1,15 +0,0 @@
-$num_instances = 2
-$vm_memory ||= 2048
-$os = "ubuntu2004"
-
-$control_plane_instances = 1
-$etcd_instances = 1
-
-# For CI we are not worried about data persistence across reboot
-$libvirt_volume_cache = "unsafe"
-
-# Checking for box update can trigger API rate limiting
-# https://www.vagrantup.com/docs/vagrant-cloud/request-limits.html
-$box_check_update = false
-
-$network_plugin = "kube-router"

tests/files/vagrant_ubuntu20-kube-router-sep.yml

@@ -1,8 +0,0 @@
----
-# Instance settings
-cloud_image: ubuntu-2004
-mode: separate
-
-# Kubespray settings
-bootstrap_os: ubuntu
-kube_network_plugin: kube-router

tests/files/vagrant_ubuntu20-kube-router-svc-proxy.rb

@@ -1,10 +0,0 @@
-$os = "ubuntu2004"
-
-# For CI we are not worried about data persistence across reboot
-$libvirt_volume_cache = "unsafe"
-
-# Checking for box update can trigger API rate limiting
-# https://www.vagrantup.com/docs/vagrant-cloud/request-limits.html
-$box_check_update = false
-
-$network_plugin = "kube-router"

tests/files/vagrant_ubuntu20-kube-router-svc-proxy.yml

@@ -1,10 +0,0 @@
----
-# Instance settings
-cloud_image: ubuntu-2004
-mode: separate
-
-# Kubespray settings
-bootstrap_os: ubuntu
-kube_network_plugin: kube-router
-
-kube_router_run_service_proxy: true

tests/scripts/create-tf.sh

@@ -1,5 +0,0 @@
-#!/bin/bash
-set -euxo pipefail
-
-cd ..
-terraform -chdir="contrib/terraform/$PROVIDER" apply -auto-approve -parallelism=1

tests/scripts/delete-tf.sh

@@ -1,5 +0,0 @@
-#!/bin/bash
-set -euxo pipefail
-
-cd ..
-terraform -chdir="contrib/terraform/$PROVIDER" destroy -auto-approve

tests/scripts/md-table/main.py

@@ -90,7 +90,7 @@ for f in files:
 
     container_manager = y.get('container_manager', 'containerd')
     network_plugin = y.get('kube_network_plugin', 'calico')
-    x = re.match(r"^[a-z-]+_([a-z0-9]+).*", f.name)
+    x = re.match(r"^([a-z-]+_)?([a-z0-9]+).*", f.name)
-    operating_system = x.group(1)
+    operating_system = x.group(2)
     data.set(container_manager=container_manager, network_plugin=network_plugin, os=operating_system)
 print(data.jinja(), file=open(args.output, 'w'))

tests/scripts/testcases_cleanup.sh

@@ -1,4 +0,0 @@
-#!/bin/bash
-set -euxo pipefail
-
-make -C tests delete-${CI_PLATFORM} -s

tests/scripts/testcases_prepare.sh

@@ -1,7 +0,0 @@
-#!/bin/bash
-set -euxo pipefail
-
-mkdir -p /.ssh
-mkdir -p cluster-dump
-mkdir -p $HOME/.ssh
-ansible-playbook --version