Was look at using Helm upgrade/install --wait for our continuous deployment but I had two concerns.

1. should waiting daemonset ignore nodes that are not ready and/or honor max unavailable. The concern is that in large node pools you might never succeed as at most points in time you at least have 1 unready node.

2. In deployments, daesmonsets and stateful sets should some sort of label be used to select only the current revisionnumber when grabbing replicaset/pods)? It seems like without it we could pick up older or newer replicasets/pods and wait on the wrong thing. That _may _ be what is happening in this issue. Maybe this is already is handled and I'm missing it.

Happy to contribute some code but wanted to see if you thought this was valid and that wait was worth investing in (maybe there's a diferent solution preferred for monitoring releases) before I went forward.