Better --check compatibility for ceph-mon role

Carefully chosen "always_run: true" parameters for read-only tasks that
register variables. This enables --check runs (at least on deployed
clusters).

roles/ceph-mon/tasks/docker/checks.yml

@@ -17,6 +17,7 @@
   changed_when: false
   failed_when: false
   register: statleftover
+  always_run: true
 
 - name: fail if we find existing cluster files
   fail:

roles/ceph-mon/tasks/docker/create_configs.yml

@@ -18,6 +18,7 @@
     removes="{{ fetch_directory }}/ceph_cluster_uuid.conf"
   changed_when: false
   register: cluster_uuid
+  always_run: true
   become: false
   when: generate_fsid
 

roles/ceph-mon/tasks/docker/dirs_permissions.yml

@@ -7,6 +7,7 @@
   failed_when: false
   run_once: true
   register: ceph_version
+  always_run: true
 
 - set_fact:
     after_hamer=True

roles/ceph-mon/tasks/docker/fetch_configs.yml

@@ -17,6 +17,7 @@
   become: false
   failed_when: false
   register: statconfig
+  always_run: true
 
 - name: try to fetch ceph config and keys
   copy:

roles/ceph-mon/tasks/docker/main.yml

@@ -4,10 +4,12 @@
   register: ceph_health
   changed_when: false
   failed_when: false
+  always_run: true
 
 - name: check if it is Atomic host
   stat: path=/run/ostree-booted
   register: stat_ostree
+  always_run: true
 
 - name: set fact for using Atomic host
   set_fact:

roles/ceph-mon/tasks/docker/selinux.yml

@@ -3,6 +3,7 @@
   command: getenforce
   register: sestatus
   changed_when: false
+  always_run: true
 
 - name: set selinux permissions
   shell: chcon -Rt svirt_sandbox_file_t {{ item }}

roles/ceph-mon/tasks/rbd_pool.yml

@@ -4,6 +4,7 @@
     ceph --connect-timeout 5 --cluster {{ cluster }} df | awk '/rbd/ {print $3}'
   changed_when: false
   failed_when: false
+  always_run: true
   register: rbd_pool_df
 
 - name: check pg num for rbd pool
@@ -11,6 +12,7 @@
     ceph --connect-timeout 5 --cluster {{ cluster }} osd pool get rbd pg_num | awk '{print $2}'
   changed_when: false
   failed_when: false
+  always_run: true
   register: rbd_pool_pgs
 
 - name: destroy and recreate rbd pool if osd_pool_default_pg_num is not honoured

roles/ceph-mon/tasks/secure_cluster.yml

@@ -2,6 +2,7 @@
 - name: collect all the pools
   command: rados --cluster {{ cluster }} lspools
   register: ceph_pools
+  always_run: true
   when: ceph_release_num.{{ ceph_release }} >= ceph_release_num.hammer
 
 - name: secure the cluster