[Update] Playbook for updating ceph should handle clusters without ceph.

[Ilya Bumarskov]

Update_ceph playbook failed on cluster without ceph:

ansible-playbook playbooks/update_ceph.yml -e '{"env_id":1,"restart_ceph":false}':

TASK [Update radosgw package] **************************************************
An exception occurred during task execution. To see the full traceback, use -vvv. The error was: KeyError: 'changed'
fatal: [node-2.test.domain.local]: FAILED! => {"changed": false, "failed": true, "module_stderr": "Traceback (most recent call last):\n File \"/tmp/ansible_EARU3s/ansible_module_apt.py\", line 909, in <module>\n main()\n File \"/tmp/ansible_EARU3s/ansible_module_apt.py\", line 892, in main\n if updated_cache and not retvals['changed']:\nKeyError: 'changed'\n", "module_stdout": "", "msg": "MODULE FAILURE"}

NO MORE HOSTS LEFT *************************************************************

PLAY RECAP *********************************************************************
node-1.test.domain.local : ok=1 changed=0 unreachable=0 failed=0
node-2.test.domain.local : ok=3 changed=1 unreachable=0 failed=1
node-3.test.domain.local : ok=1 changed=0 unreachable=0 failed=0

[Alexey Stupnikov (astupnikov)]

This bug is fixed by https://review.fuel-infra.org/#/c/29067/ patch.

[Ilya Bumarskov (ibumarskov)]

Was verified on snapshot-id #631