A new cluster with the same name is not created during deletion of old one

[olegvn88]

A new cluster with the same name is not created during deletion of old one. The previous one works with 0 workers.

Preconditions:
At least one cluster should be created.

Steps to reproduce:

To delete cluster:

1. Open the Oshinko Spark Cluster Management home page.
2. Click on the "Actions" button.
3. Click on the "Delete Cluster" button, and click on the "Delete" button in the opened window window and immediately go to the next step.

To create a new one cluster with the same name:

4. Click on "Deploy" button.
5. Input Name of the cluster.
6. Input the "Number of workers" value equals 2.
7. Click on "Deploy" button.

Actual result: We've got an error message "object is being deleted: deploymentconfigs "sparkcluster-w" already exists", the old one cluster runs with 0 workers

Expected result: A new cluster must be created with 2 workers.

[crobby]

I think the actual behaviour we want here is to not be able to create a new cluster with the same name as an existing one until all of the original cluster is gone away. I think that would be consistent with typical behaviour in openshift. The webui should, before trying to create the new cluster, check to be sure that all of the objects needed will be able to be created without a name collision. If there is a naming collision, we should pass that back to the user. They can either wait or try a different name.