Description of the problem: 
When a workflow instance is aborted, related generated processes that take up cluster resources should stop running As well. For example, the MTA 7 workflow generates a related task pod in the openshift-mta namespace when it runs.  Aborting the workflow should terminate the task pod. Currently, when the workflow is aborted, the related task pod keeps on running and taking up cluster resources. 

How reproducible: 100%

Steps to reproduce:

1. Run a workflow like MTA 7 v0.2.33

2. Run MTA 7 workflow, and abort after 30 seconds

3. Look at pod statuses in openshift-mta namespace

Actual results: Task pod continues to be in Running state, even though the workflow instance that spawned it was aborted 

Expected results: Task pod should be terminated.