teardown logic should only execute when setup is successful about kubetest HOT 1 CLOSED

Well this became a bit of a rabbit hole.

I believe this issue is a bit different than I originally described. I don't think that we want to catch a setup failure and limit teardown based on that because there could be a partial failure in setup where some resources are already created, so skipping teardown would leave them on the cluster.

Instead, I think we should check whether we can connect to the cluster at all, and if so attempt to clean up resources as we do now. If not, then just raise a warning. There are probably smarter ways to go about this which we can look into later, but getting something in place that works first would be nice.

It would also be nice to be able to check if we are already configured to use a cluster -- if so, we could continue with teardown, but if we failed before ever getting to load cluster configs, then there is no way we could have loaded stuff onto the cluster. Unfortunately, it doesn't seem like there is a great way to do this with the kubernetes python api. There are ways to do it.. but they are not great because there would be some assumptions being made about cluster configuration which may not always be true.

from kubetest.