Alexander Brand

Thanks @omazilov. We don't have it in the short term roadmap, but might get to it in the medium/long term.

Discussion around this (including that specific label) has been ongoing for some time now: https://github.com/kubernetes/kubernetes/pull/39112. I think we can implement this once there's consensus, but from what I can see...

Is there a reason for waiting until the smoke test resources are cleaned up? Adding that will only increase the installation time, and I am not sure I see the...

This is expected to happen as we are currently not waiting until the deployments are completely gone. I would argue we shouldn't wait, given that we are not smoke testing...

The namespace idea is probably the best, but I would avoid waiting until all resources are cleaned up. All that is going to do is slow down the overall installation...

What are your thoughts on ReplicaSet + Anti Affinity? This would allow different users to choose how wide they want to go when scaling KubeDNS. I am not sure if...

Another potential option is to inspect the output of the `terraform plan` command, and determine whether terraform is planning to destroy infrastructure. I dislike the parsing of arbitrary output, as...

I am now wondering if we could use terraform as a library to ingest the plan, and determine if there are any resources marked for destruction.

In my mind, the effort here is to automate the ~100 node test so that we can run it on schedule. This will also help in reducing the manual process...

The contiv workflow does not have a separate validation step because of the way it needs to be installed. The validation is performed in the same play as the installation,...