In the scenario that tripleo is connecting to an external or per-installed ceph cluster *and* there isn't a viable ceph container image (in my case I'm on ppc64le) we need ceph-common[1]
Without ceph-common the deploy wiull fail with something like:
---
2018-08-31 01:18:02,587 p=8101 u=mistral | TASK [ceph-common : get ceph version] ******************************************
2018-08-31 01:18:02,587 p=8101 u=mistral | Friday 31 August 2018 01:18:02 -0400 (0:00:00.044) 0:00:17.102 *********
2018-08-31 01:18:02,900 p=8101 u=mistral | fatal: [192.168.0.67]: FAILED! => {"changed": false, "cmd": "ceph --version", "failed": true, "msg": "[Errno 2] No such file or directory", "rc": 2}
2018-08-31 01:18:02,963 p=8101 u=mistral | fatal: [192.168.0.76]: FAILED! => {"changed": false, "cmd": "ceph --version", "failed": true, "msg": "[Errno 2] No such file or directory", "rc": 2}
2018-08-31 01:18:02,978 p=8101 u=mistral | fatal: [192.168.0.71]: FAILED! => {"changed": false, "cmd": "ceph --version", "failed": true, "msg": "[Errno 2] No such file or directory", "rc": 2}
---
[1] http://mirror.centos.org/altarch/7/storage/ppc64le/ceph-luminous/ceph-common-12.2.5-0.el7.ppc64le.rpm
Related fix proposed to branch: master /review. openstack. org/599915
Review: https:/