Allow me to explain this clearly.
This bug will not be introduced while provisioning a new cluster.
The problem here is that for an existing running cluster, you have oozie db running when you try to apply some changes like configurations or components by running puppet apply again. At that moment you'll get an error shows that Exec[Oozie DB init] failed, which will be sort of confusing since we don't know what if there is actually an error occurred in my environment or not.
Here's the reproduce steps using vagrant-puppet:
1.) specify oozie to be deployed in the cluster
2.) Spin up the cluster, which will succeed
3.) run puppet apply again
$ vagrant provision --provision-with puppet
This will introduce the problem.