RHCS2 - ceph-ansible - rolling_update fails at TASK [waiting for clean pgs...]

Issue

We are updateing out RHCS 2.x environment with ceph-ansible-3.0.25-1 version of rolling_update.yml and the ansible playbook fails at TASK [waiting for clean pgs...] even when ceph is HEALTH_OK:

# ceph -s
    cluster f88975b3-3e89-4e9d-8d6e-8b4671a2f496
     health HEALTH_OK
     monmap e2: 3 mons at {ceph1=10.37.192.104:6789/0,ceph2=10.37.192.105:6789/0,ceph3=10.37.192.106:6789/0}
            election epoch 58, quorum 0,1,2 ceph1,ceph2,ceph3
     osdmap e1271: 5 osds: 5 up, 5 in
            flags sortbitwise,require_jewel_osds,recovery_deletes
      pgmap v590267: 176 pgs, 8 pools, 1108 MB data, 15153 objects
            3741 MB used, 1377 GB / 1381 GB avail
                 167 active+clean
                   5 active+clean+scrubbing+deep
                   4 active+clean+scrubbing