Fix broken CI jobs (#1854)

* Fix broken CI jobs

Adjust image and image_family scenarios for debian.
Checkout CI file for upgrades

* add debugging to file download

* Fix download for alternate playbooks

* Update ansible ssh args to force ssh user

* Update sync_container.yml

.gitlab-ci.yml

@@ -96,6 +96,8 @@ before_script:
     # Uncomment when gitlab kargo repo has tags
     #- test "${UPGRADE_TEST}" != "false" && git fetch --all && git checkout $(git describe --tags $(git rev-list --tags --max-count=1))
     - test "${UPGRADE_TEST}" != "false" && git checkout 72ae7638bcc94c66afa8620dfa4ad9a9249327ea
+    # Checkout the CI vars file so it is available
+    - test "${UPGRADE_TEST}" != "false" && git checkout "${CI_BUILD_REF}" tests/files/${CI_JOB_NAME}.yml
 
 
     # Create cluster

ansible.cfg

@@ -1,7 +1,7 @@
 [ssh_connection]
 pipelining=True
-ssh_args = -o ControlMaster=auto -o ControlPersist=30m -o ConnectionAttempts=100
+ansible_ssh_common_args = -o ControlMaster=auto -o ControlPersist=30m -o ConnectionAttempts=100
-#ssh_args = -F ./ssh-bastion.conf -o ControlMaster=auto -o ControlPersist=30m -o ConnectionAttempts=100
+#ansible_ssh_common_args = -F {{ inventory_dir|quote }}/ssh-bastion.conf -o ControlMaster=auto -o ControlPersist=30m -o ConnectionAttempts=100
 #control_path = ~/.ssh/ansible-%%r@%%h:%%p
 [defaults]
 host_key_checking=False

extra_playbooks/upgrade-only-k8s.yml

@@ -3,6 +3,7 @@
 ### * Will not upgrade etcd
 ### * Will not upgrade network plugins
 ### * Will not upgrade Docker
+### * Will not pre-download containers or kubeadm
 ### * Currently does not support Vault deployment.
 ###
 ### In most cases, you probably want to use upgrade-cluster.yml playbook and

roles/download/tasks/download_file.yml

@@ -1,4 +1,10 @@
 ---
+- name: file_download | Downloading...
+  debug:
+    msg:
+      - "URL: {{ download.url }}"
+      - "Dest: {{ download.dest }}"
+
 - name: file_download | Create dest directory
   file:
     path: "{{local_release_dir}}/{{download.dest|dirname}}"

roles/download/tasks/sync_container.yml

@@ -62,7 +62,7 @@
   synchronize:
     src: "{{ fname }}"
     dest: "{{ fname }}"
-    use_ssh_args: yes
+    use_ssh_args: "{{ has_bastion | default(false) }}"
     mode: pull
   delegate_to: localhost
   delegate_facts: no
@@ -81,7 +81,7 @@
   synchronize:
     src: "{{ fname }}"
     dest: "{{ fname }}"
-    use_ssh_args: yes
+    use_ssh_args: "{{ has_bastion | default(false) }}"
     mode: push
   delegate_to: localhost
   delegate_facts: no

tests/ansible.cfg

@@ -1,6 +1,6 @@
 [ssh_connection]
 pipelining=True
-ssh_args = -o ControlMaster=auto -o ControlPersist=30m -o ConnectionAttempts=100
+ansible_ssh_common_args = -o ControlMaster=auto -o ControlPersist=30m -o ConnectionAttempts=100
 [defaults]
 forks = 20
 host_key_checking=False

tests/cloud_playbooks/create-gce.yml

@@ -29,7 +29,8 @@
       gce:
         instance_names: "{{instance_names}}"
         machine_type: "{{ cloud_machine_type }}"
-        image_family: "{{ cloud_image }}"
+        image: "{{ cloud_image | default(omit) }}"
+        image_family: "{{ cloud_image_family | default(omit) }}"
         preemptible: "{{ preemptible }}"
         service_account_email: "{{ gce_service_account_email }}"
         pem_file: "{{ gce_pem_file | default(omit)}}"

tests/files/centos-weave-kubeadm.yml

@@ -1,5 +1,5 @@
 # Instance settings
-cloud_image: centos-7
+cloud_image_family: centos-7
 cloud_machine_type: "n1-standard-1"
 cloud_region: us-central1-b
 mode: ha
@@ -9,7 +9,7 @@ startup_script: ""
 kube_network_plugin: weave
 weave_cpu_limit: "100m"
 weave_cpu_requests: "100m"
-kubeadm_enabled: "true"
+kubeadm_enabled: true
 deploy_netchecker: true
 kubedns_min_replicas: 1
 cloud_provider: gce

tests/files/centos7-calico-ha.yml

@@ -1,12 +1,12 @@
 # Instance settings
-cloud_image: centos-7
+cloud_image_family: centos-7
 cloud_region: europe-west1-b
 mode: ha-scale
 
 # Deployment settings
 kube_network_plugin: calico
-download_localhost: "true"
+download_localhost: true
-download_run_once: "true"
+download_run_once: true
 deploy_netchecker: true
 kubedns_min_replicas: 1
 cloud_provider: gce

tests/files/centos7-flannel-addons.yml

@@ -1,5 +1,5 @@
 # Instance settings
-cloud_image: centos-7
+cloud_image_family: centos-7
 cloud_region: us-west1-a
 cloud_machine_type: "n1-standard-1"
 mode: default

tests/files/coreos-alpha-weave-ha.yml

@@ -1,5 +1,5 @@
 # Instance settings
-cloud_image: coreos-alpha
+cloud_image_family: coreos-alpha
 cloud_region: us-west1-a
 mode: ha-scale
 startup_script: 'systemctl disable locksmithd && systemctl stop locksmithd'

tests/files/coreos-calico-aio.yml

@@ -1,5 +1,5 @@
 # Instance settings
-cloud_image: coreos-stable
+cloud_image_family: coreos-stable
 cloud_region: us-west1-b
 cloud_machine_type: "n1-standard-2"
 mode: aio

tests/files/coreos-canal.yml

@@ -1,5 +1,5 @@
 # Instance settings
-cloud_image: coreos-stable
+cloud_image_family: coreos-stable
 cloud_region: us-east1-b
 mode: default
 startup_script: 'systemctl disable locksmithd && systemctl stop locksmithd'

tests/files/rhel7-canal-sep.yml

@@ -1,5 +1,5 @@
 # Instance settings
-cloud_image: rhel-7
+cloud_image_family: rhel-7
 cloud_region: us-east1-b
 mode: separate
 

tests/files/rhel7-weave.yml

@@ -1,5 +1,5 @@
 # Instance settings
-cloud_image: rhel-7
+cloud_image_family: rhel-7
 cloud_region: europe-west1-b
 mode: default
 

tests/files/ubuntu-canal-ha.yml

@@ -1,5 +1,5 @@
 # Instance settings
-cloud_image: ubuntu-1604-lts
+cloud_image_family: ubuntu-1604-lts
 cloud_region: europe-west1-b
 mode: ha
 

tests/files/ubuntu-canal-kubeadm.yml

@@ -1,5 +1,5 @@
 # Instance settings
-cloud_image:  ubuntu-1604-lts
+cloud_image_family:  ubuntu-1604-lts
 cloud_machine_type: "n1-standard-1"
 cloud_region: europe-west1-b
 mode: ha
@@ -7,7 +7,7 @@ mode: ha
 # Deployment settings
 bootstrap_os: ubuntu
 kube_network_plugin: canal
-kubeadm_enabled: "true"
+kubeadm_enabled: true
 deploy_netchecker: true
 kubedns_min_replicas: 1
 cloud_provider: gce

tests/files/ubuntu-flannel-sep.yml

@@ -1,5 +1,5 @@
 # Instance settings
-cloud_image: ubuntu-1604-lts
+cloud_image_family: ubuntu-1604-lts
 cloud_region: europe-west1-b
 mode: separate
 

tests/files/ubuntu-rkt-sep.yml

@@ -1,5 +1,5 @@
 # Instance settings
-cloud_image: ubuntu-1604-lts
+cloud_image_family: ubuntu-1604-lts
 cloud_region: us-central1-b
 mode: separate
 
@@ -8,6 +8,8 @@ bootstrap_os: ubuntu
 kube_network_plugin: flannel
 etcd_deployment: rkt
 kubelet_deployment: rkt
+download_localhost: true
+download_run_once: true
 deploy_netchecker: true
 kubedns_min_replicas: 1
 cloud_provider: gce

tests/files/ubuntu-vault-sep.yml

@@ -1,6 +1,6 @@
 # Instance settings
 cloud_machine_type: "n1-standard-1"
-cloud_image: ubuntu-1604-lts
+cloud_image_family: ubuntu-1604-lts
 cloud_region: us-central1-b
 mode: separate
 

tests/files/ubuntu-weave-sep.yml

@@ -1,5 +1,5 @@
 # Instance settings
-cloud_image: ubuntu-1604-lts
+cloud_image_family: ubuntu-1604-lts
 cloud_region: us-central1-b
 mode: separate
 

upgrade-cluster.yml

@@ -32,6 +32,7 @@
     - role: rkt
       tags: rkt
       when: "'rkt' in [etcd_deployment_type, kubelet_deployment_type, vault_deployment_type]"
+    - { role: download, tags: download, skip_downloads: false }
 
 - hosts: etcd:k8s-cluster:vault
   any_errors_fatal: "{{ any_errors_fatal | default(true) }}"