Failover does not work if gluster cluster with cifs smb access was created with gdeploy

Issue

• There is a bug in gdeploy while it creates gluster cluster with smb/cifs access. The dedicated user for samba while defined in gdeploy config file is being created only on first node in the cluster. This is causing issue while CTDB failover takes place - if any security is enabled for samba, remount to new node after failover will fail.