Created attachment 1322590 [details]
File contains contents of inventory file and ansible-playbook log

Description of problem:
Restart ceph-defaults: retsceph mds task is failing on a node before configuring mdss. ceph-ansible was configured to colocate containerized osds with containerized mdss. While running ceph-defaults handlers after configuring osds, restart ceph mdss task failed as mds was not yet configured on the node. Same task was skipped on other nodes. But it should have skipped on this node also as mds was not yet configured.

Version-Release number of selected component (if applicable):
ceph-ansible-3.0.0-0.1.rc4.el7cp.noarch


Steps to Reproduce:
1. Configure ceph-ansible to get a cluster with osds container colocated with mds.
2. Run ansible-playbook site-docker.yml

Actual results:
RUNNING HANDLER [ceph-defaults : restart ceph mdss] **********************************************************************************************
failed: [magna025 -> magna025] (item=magna025) => {"failed": true, "item": "magna025", "msg": "Could not find the requested service ceph-mds@magna025: host"}


Expected results:
restart ceph mdss must have been skipped before mds is configured

Additional info:
$ cat /etc/ansible/hosts 
--------
[osds]
magna025
-------
[mdss]
magna025