SM provision of 16.04.2 cluster stuck due to keystone service not found

R4.0-23 mitaka SM provisioning 16.04.2 newton cluster. The problem was not seen in R4.0-22.

root@servermanager:~# server-manager status server

+---------+-------------------------------+------------+-------------------+
| id | status | ip_address | mac_address |
+---------+-------------------------------+------------+-------------------+
| server4 | provision_issued | 10.0.0.7 | 02:BA:39:30:B2:78 |
| server3 | post_exec_vnc_galera_started | 10.0.0.6 | 02:AB:8C:A3:D1:B0 |
| server5 | provision_issued | 10.0.0.8 | 02:32:59:0E:EA:95 |
| server8 | provision_issued | 10.0.0.11 | 02:77:E9:47:A7:9D |
| server9 | provision_issued | 10.0.0.12 | 02:A9:13:3D:22:CD |
| server2 | pre_exec_vnc_galera_completed | 10.0.0.5 | 02:F8:AA:1E:83:AA |
| server7 | provision_issued | 10.0.0.10 | 02:65:53:EF:E2:18 |
| server6 | provision_issued | 10.0.0.9 | 02:D9:18:78:B4:3E |
| server1 | pre_exec_vnc_galera_completed | 10.0.0.4 | 02:CE:9B:D8:C6:4B |
+---------+-------------------------------+------------+-------------------+
root@servermanager:~#

            "provision_role_sequence": "{'completed': [(u'server3', 'keepalived', '2017_07_06__07_24_05'), (u'server1', 'keepalived', '2017_07_06__07_24_07'), (u'server2', 'keepalived', '2017_07_06__07_24_07'), (u'server3', 'haproxy', '2017_07_06__07_24_37'), (u'server1', 'haproxy', '2017_07_06__07_24_39'), (u'server2', 'haproxy', '2017_07_06__07_24_40'), (u'server1', 'openstack', '2017_07_06__07_32_10'), (u'server2', 'openstack', '2017_07_06__07_32_31'), (u'server3', 'openstack', '2017_07_06__07_37_20'), (u'server3', 'pre_exec_vnc_galera', '2017_07_06__07_38_48'), (u'server2', 'pre_exec_vnc_galera', '2017_07_06__07_39_57'), (u'server1', 'pre_exec_vnc_galera', '2017_07_06__07_41_23')], 'steps': [[(u'server3', u'post_exec_vnc_galera')], [(u'server2', u'post_exec_vnc_galera')], [(u'server1', u'post_exec_vnc_galera')], [(u'server1', 'post_provision'), (u'server2', 'post_provision'), (u'server3', 'post_provision')]]}",

Jul 6 15:56:53 server3 puppet-agent[5310]: (/Stage[ha_config]/Contrail::Ha_config/Exec[exec-transfer-keys]/returns) Failed to restart keystone.service: Unit keystone.service not found.
Jul 6 15:56:53 server3 puppet-agent[5310]: python /opt/contrail/bin/transfer_keys.py 10.10.0.6 "/etc/ssl/" root c0ntrail123 && echo exec-transfer-keys >> /etc/contrail/contrail_openstack_exec.out returned 1 instead of one of [0]
Jul 6 15:56:53 server3 puppet-agent[5310]: (/Stage[ha_config]/Contrail::Ha_config/Exec[exec-transfer-keys]/returns) change from notrun to 0 failed: python /opt/contrail/bin/transfer_keys.py 10.10.0.6 "/etc/ssl/" root c0ntrail123 && echo exec-transfer-keys >> /etc/contrail/contrail_openstack_exec.out returned 1 instead of one of [0]
Jul 6 15:56:53 server3 puppet-agent[5310]: (/Stage[ha_config]/Contrail::Ha_config/Contrail::Lib::Report_status[post_exec_vnc_galera_completed]/Exec[contrail-status-post_exec_vnc_galera_completed]) Dependency Exec[exec-transfer-keys] has failures: true
Jul 6 15:56:53 server3 puppet-agent[5310]: (/Stage[ha_config]/Contrail::Ha_config/Contrail::Lib::Report_status[post_exec_vnc_galera_completed]/Exec[contrail-status-post_exec_vnc_galera_completed]) Skipping because of failed dependencies