Look into Failure for Cluster Teardown

[kensipe]

A KEP PR resulting in a test failure... it is a failure for waiting for teardown. We should understand if this failure is a concern and consider removing it as a test result if we do not care.

harness.go:579: fatal error running commands: command "./bin/kubectl-kudo init --unsafe-self-signed-webhook-ca --kudo-image kudobuilder/controller:test --kudo-image-pull-policy Never --wait" exceeded 300 sec timeout

From: https://github.com/kudobuilder/kudo/pull/1794

What would you like to be added:

Why is this needed:

[asekretenko]

@kensipe Is this really a teardown failure?..

From what I could see in the logs, this is a failure of waiting for kudo to be installed, and it looks like kudo controller manager pod has never been scheduled in this run.

Something very similar is observed in upgrade tests on a much more regular basis, see https://github.com/kudobuilder/kudo/issues/1783