ceph-handler: Fix osd handler in check mode

Run the Ceph commands that only gather information (without making any changes
to the cluster) when running Ansible in check mode.

This allows the tasks that depend on the variables set by those tasks to
succeed in check mode.

Signed-off-by: Benoît Knecht <bknecht@protonmail.ch>
(cherry picked from commit 498acd7)

roles/ceph-handler/tasks/handler_osds.yml

@@ -15,13 +15,15 @@
       delegate_to: "{{ groups.get(mon_group_name, [])[0] }}"
       run_once: true
       changed_when: false
+      check_mode: false
 
     - name: get balancer module status
       command: "{{ ceph_cmd }} --cluster {{ cluster }} balancer status -f json"
       register: balancer_status
       run_once: true
       delegate_to: "{{ groups[mon_group_name][0] }}"
       changed_when: false
+      check_mode: false
 
     - name: set_fact pools_pgautoscaler_mode
       set_fact: