Cleanup deployments from namespace

[jayz22]

Why

Resolves #347

The clean command doesn't clean up the deployments (currently only used in parallel catchup v2).
Which results to the v2 mission isn't cleanup thoroughly in certain conditions :

• if an entire supercluster mission is aborted (pressing the abort button in Jenkins)
• and the abort doesn't wait enough time for v2 graceful cleanup to finish (it takes longer due to trying to copy/save logs from all pods)
• it force kills the v2 mission (with 3 consecutive SIGTERM)
• then it relies on clean command to clean up everything, which will have some left over. This prevents a new v2 mission from being spawn.

This ensure all relevant k8s artifacts gets cleaned up regardless.