Run containerized kraken as a kubernetes/openshift deployment about krkn HOT 4 CLOSED

When we run Kraken/Cerberus as a Kubernetes/OpenShift object, there might be cases where the pod which is running the tool might get disrupted unless we add a way to blacklist it during the run. For instance, we might not get the results from the pod running Kraken or Cerberus when the API iteself is down. When we run them external to OpenShift, we can get a signal and catch test failures even when the API is down. Thoughts?

But agree that we might need to support both - running it internal/external to the cluster to improve the ease of use and users can pick the option depending on the use case.

from krkn.

We can do something similar to https://github.com/openshift-scale/cerberus/tree/master/containers keeping in mind to avoid killing anything in namespace/node with kraken running.

from krkn.

@paigerube14 do you want to pick this?

from krkn.

Fixed by #24. Thanks @paigerube14.

from krkn.