rolling_update: fix get fsid for containers

When running ansible2.4-update_docker_cluster there is an issue on the
"get current fsid" task. The current task only works for
non-containerized deployment but will run all the time (even for
containerized). This currently results in the following error:

TASK [get current fsid] ********************************************************
task path: /home/jenkins-build/build/workspace/ceph-ansible-prs-luminous-ansible2.4-update_docker_cluster/rolling_update.yml:214
Tuesday 22 May 2018  22:48:32 +0000 (0:00:02.615)       0:11:01.035 ***********
fatal: [mgr0 -> mon0]: FAILED! => {
    "changed": true,
    "cmd": [
        "ceph",
        "--cluster",
        "test",
        "fsid"
    ],
    "delta": "0:05:00.260674",
    "end": "2018-05-22 22:53:34.555743",
    "rc": 1,
    "start": "2018-05-22 22:48:34.295069"
}

STDERR:

2018-05-22 22:48:34.495651 7f89482c6700  0 -- 192.168.17.10:0/1022712 >> 192.168.17.12:6789/0 pipe(0x7f8944067010 sd=4 :42654 s=1 pgs=0 cs=0 l=1 c=0x7f894405d510).connect protocol feature mismatch, my 83ffffffffffff < peer 481dff8eea4fffb missing 400000000000000
2018-05-22 22:48:34.495684 7f89482c6700  0 -- 192.168.17.10:0/1022712 >> 192.168.17.12:6789/0 pipe(0x7f8944067010 sd=4 :42654 s=1 pgs=0 cs=0 l=1 c=0x7f894405d510).fault

This is not really representative on the real error since the 'ceph' cli is available on that machine.
On other environments we will have something like "command not found: ceph".

Signed-off-by: Sébastien Han <seb@redhat.com>

infrastructure-playbooks/rolling_update.yml

@@ -211,10 +211,23 @@
   become: True
 
   pre_tasks:
-    - name: get current fsid
+    - name: non container - get current fsid
       command: "ceph --cluster {{ cluster }} fsid"
-      register: cluster_uuid
+      register: cluster_uuid_non_container
       delegate_to: "{{ groups[mon_group_name][0] }}"
+      when:
+        - not containerized_deployment
+
+    - name: container - get current fsid
+      command: "docker exec ceph-mon-{{ hostvars[groups[mon_group_name][0]]['ansible_hostname'] }} ceph --cluster {{ cluster }} fsid"
+      register: cluster_uuid_container
+      delegate_to: "{{ groups[mon_group_name][0] }}"
+      when:
+        - containerized_deployment
+
+    - name: set_fact ceph_cluster_fsid
+      set_fact:
+        ceph_cluster_fsid: "{{ cluster_uuid_container.stdout if containerized_deployment else cluster_uuid_non_container.stdout }}"
 
     - name: non container | create ceph mgr keyring(s)
       command: "ceph --cluster {{ cluster }} auth get-or-create mgr.{{ hostvars[item]['ansible_hostname'] }} mon 'allow profile mgr' osd 'allow *' mds 'allow *' -o /etc/ceph/{{ cluster }}.mgr.{{ hostvars[item]['ansible_hostname'] }}.keyring"
@@ -243,7 +256,7 @@
     - name: fetch ceph mgr key(s)
       fetch:
         src: "{{ ceph_conf_key_directory }}/{{ cluster }}.mgr.{{ hostvars[item]['ansible_hostname'] }}.keyring"
-        dest: "{{ fetch_directory }}/{{ fsid }}/{{ ceph_conf_key_directory }}/"
+        dest: "{{ fetch_directory }}/{{ ceph_cluster_fsid }}/{{ ceph_conf_key_directory }}/"
         flat: yes
         fail_on_missing: no
       delegate_to: "{{ groups[mon_group_name][0] }}"