fix(tests,kubevirt_vm): Make waiting for VM more robust

This changes the runme.sh of the kubevirt_vm integration tests to retry
the connection to the VM until a login is possible. This is necessary
since it is not possible with Ansible alone to retry a task in case the
connection failed with the unreachable status. This can happen when the
sshd of the VM already accepts connections but a login is not yet
possible because the VM is still booting up.

Signed-off-by: Felix Matouschek <fmatouschek@redhat.com>

tests/integration/targets/kubevirt_vm/verify.yml

@@ -1,27 +1,4 @@
 ---
-- name: Wait for SSH
-  connection: local
-  gather_facts: false
-  hosts: localhost
-  tasks:
-    - name: Wait up to 900 seconds for port 22 to become open
-      ansible.builtin.wait_for:
-        port: 22
-        host: "{{ hostvars['default-testvm'].ansible_host }}"
-        delay: 30
-        timeout: 900
-
-- name: Connect to VM
-  gather_facts: true
-  hosts: default-testvm
-  remote_user: cloud-user
-  vars:
-    ansible_python_interpreter: /usr/bin/python3
-  tasks:
-    - name: Print VM facts
-      ansible.builtin.debug:
-        var: ansible_facts
-
 - name: Verify creation with existing VM
   connection: local
   gather_facts: false