Don't have interface with glusterfs stroage generated after provisioning kubevirt-apb via web console

[qwang1]

Hi there,

So far the latest kubevirt-apb version is v0.4.1-alpha.2. I tried to provision it with glusterfs storage checked on OCP v3.10.0-0.47.0 + OpenStack. There is no kubvirt storageclass generated. Seems it didn't run storage part. If I use ansible-playbook to deploy kubevirt.yml, this problem won't happen http://pastebin.test.redhat.com/598551.

[root@cnv-executor-qwang-apb-master1 ~]# oc get all 
NAME                                   READY     STATUS    RESTARTS   AGE
pod/virt-api-59f4bdd6-5z8gg            1/1       Running   0          45s
pod/virt-api-59f4bdd6-cvc8b            1/1       Running   1          45s
pod/virt-controller-6756fcdcc9-2nc5s   1/1       Running   0          45s
pod/virt-controller-6756fcdcc9-54n6v   0/1       Running   0          45s

NAME               TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)   AGE
service/virt-api   ClusterIP   172.30.122.46   <none>        443/TCP   45s

NAME                                DESIRED   CURRENT   READY     UP-TO-DATE   AVAILABLE   NODE SELECTOR   AGE
daemonset.extensions/virt-handler   0         0         0         0            0           <none>          45s

NAME                                    DESIRED   CURRENT   UP-TO-DATE   AVAILABLE   AGE
deployment.extensions/virt-api          2         2         2            2           45s
deployment.extensions/virt-controller   2         2         2            1           45s

NAME                                               DESIRED   CURRENT   READY     AGE
replicaset.extensions/virt-api-59f4bdd6            2         2         2         45s
replicaset.extensions/virt-controller-6756fcdcc9   2         2         1         45s

[root@cnv-executor-qwang-apb-master1 ~]# oc get storageclass
NAME                          PROVISIONER               AGE
glusterfs-storage (default)   kubernetes.io/glusterfs   2d

[root@cnv-executor-qwang-apb-master1 ~]# oc get all -n dh-virtualization-prov-s5t4s
NAME                                           READY     STATUS      RESTARTS   AGE
pod/apb-e86b7acc-7300-4783-8c93-dcba82c906ed   0/1       Completed   0          1m

[root@cnv-executor-qwang-apb-master1 ~]# oc logs pod/apb-e86b7acc-7300-4783-8c93-dcba82c906ed -n dh-virtualization-prov-s5t4s

PLAY [Provision KubeVirt] ******************************************************

TASK [ansible.kubernetes-modules : Install latest openshift client] ************
skipping: [localhost]

TASK [ansibleplaybookbundle.asb-modules : debug] *******************************
skipping: [localhost]

PLAY [localhost] ***************************************************************

TASK [kubevirt : include_tasks] ************************************************
included: /etc/ansible/roles/kubevirt-ansible/roles/kubevirt/tasks/provision.yml for localhost

TASK [kubevirt : Login As Super User] ******************************************
changed: [localhost]

TASK [kubevirt : Check if kubevirt-apb-2 exists] *******************************
changed: [localhost]

TASK [kubevirt : Create kubevirt-apb-2 namespace] ******************************
skipping: [localhost]

TASK [kubevirt : Add Privileged Policy] ****************************************
changed: [localhost] => (item=kubevirt-privileged)
changed: [localhost] => (item=kubevirt-controller)
changed: [localhost] => (item=kubevirt-infra)

TASK [kubevirt : Add Hostmount-anyuid Policy] **********************************
changed: [localhost]

TASK [kubevirt : Check for kubevirt.yaml template in /etc/ansible/roles/kubevirt-ansible/roles/kubevirt/templates] ***
ok: [localhost]

TASK [kubevirt : Check for offline v0.4.1-alpha.2 templates in /opt/apb/kubevirt-templates] ***
ok: [localhost]

TASK [kubevirt : Download KubeVirt Template] ***********************************
skipping: [localhost]

TASK [kubevirt : Copy offline templates to /tmp] *******************************
changed: [localhost]

TASK [kubevirt : Render KubeVirt Yaml] *****************************************
changed: [localhost]

TASK [kubevirt : Render BYO template] ******************************************
skipping: [localhost]

TASK [kubevirt : Create KubeVirt Resources] ************************************
changed: [localhost]

TASK [kubevirt : Check for vm templates in /etc/ansible/roles/kubevirt-ansible/roles/kubevirt/templates] ***
ok: [localhost] => (item=vm-template-fedora)
ok: [localhost] => (item=vm-template-windows2012r2)
ok: [localhost] => (item=vm-template-rhel7)

TASK [kubevirt : Copy VM templates to /tmp] ************************************

TASK [kubevirt : Check for vm templates in /opt/apb/kubevirt-templates] ********
ok: [localhost] => (item=vm-template-fedora)
ok: [localhost] => (item=vm-template-windows2012r2)
ok: [localhost] => (item=vm-template-rhel7)

TASK [kubevirt : Copy VM templates to /tmp] ************************************
 [WARNING]: when statements should not include jinja2 templating delimiters
such as {{ }} or {% %}. Found: {{ offline_vm_templates.results |
selectattr('stat.exists') | map(attribute='item') | list | length > 0 }}
changed: [localhost] => (item=vm-template-fedora)
changed: [localhost] => (item=vm-template-windows2012r2)
changed: [localhost] => (item=vm-template-rhel7)

TASK [kubevirt : Download KubeVirt default VM templates] ***********************
 [WARNING]: when statements should not include jinja2 templating delimiters
such as {{ }} or {% %}. Found: cluster == "openshift" and "{{
byo_vm_templates.results | selectattr('stat.exists') | map(attribute='item') |
list | length == 0 }}" and "{{ offline_vm_templates.results |
selectattr('stat.exists') | map(attribute='item') | list | length == 0 }}"
ok: [localhost] => (item=vm-template-fedora)
ok: [localhost] => (item=vm-template-windows2012r2)
ok: [localhost] => (item=vm-template-rhel7)

TASK [kubevirt : Create default VM templates in OpenShift Namespace] ***********
changed: [localhost] => (item=vm-template-fedora)
changed: [localhost] => (item=vm-template-windows2012r2)
changed: [localhost] => (item=vm-template-rhel7)
 [WARNING]: Could not match supplied host pattern, ignoring: masters

PLAY [masters[0]] **************************************************************
skipping: no hosts matched
 [WARNING]: Could not match supplied host pattern, ignoring: nodes

PLAY [masters nodes] ***********************************************************
skipping: no hosts matched

PLAY RECAP *********************************************************************
localhost                  : ok=15   changed=9    unreachable=0    failed=0

@rthallisey Could you take a look at it? Thanks.

[lukas-bednar]

@qwang1 This comment is not related to your issue, it is just a note about kubevirt version: the latest released is v0.5.0, and pre-released is v0.5.1-alpha.2 ... unfortunately you can not deploy these from WebConsole (because hardcoded enum in apb), but you can create serviceinstance manually with what ever params you like, for example for latest kubevirt use:

# kubevirt-apb.yml
---
apiVersion: servicecatalog.k8s.io/v1beta1
kind: ServiceInstance
metadata:
  name: kubevirt
  namespace: kube-system
spec:
  clusterServiceClassExternalName: "dh-virtualization"
  clusterServicePlanExternalName: gluster
  parameters:
    admin_user: "test_admin"
    admin_password: "123456"
    version: "0.5.1-alpha.2"

oc create -f kubevirt-apb.yml

[qwang1]

@lukas-bednar Yes, the version on apb.yml file should be updated. However, I saw a new file download-templates.sh was added and get the latest version with kubevirt release, I have no idea where this script is used.

[rthallisey]

ansibleplaybookbundle/apb-base#31 fixed the issue. I just updated the dockerhub image docker.io/ansibleplaybookbundle/kubevirt-apb:latest to make sure it has the fix. Make sure you're using newest image. Reopen if the issue persists.