Florian Ruynat

> > Do you know why does Kubespary use Nerdctl with native > > Idk, perhaps misunderstanding of "native" Agree, the `native` value seems like a mistake and not really...

Well I agree that specifying namespace could be an added option but I'm strongly against removing all namespaces as it's pretty much my daily use case :)

I have some use cases with resources handle by operator on our end, which create some pods in each namespace declared by users. And sometime I want to check all...

> @ssoriche @floryut How about setting an upper limit on the number of pods or containers to handle? When the limit is reached, stern is terminated. This also makes the...

Yes hard to set a number but something like 10/20 is still manageable (when someone know what is he doing)

Let's see what @superbrothers and @rkmathi thinks, but I'd say that it's freaking cool 👍

> We have a user 'input' without timeout and that happens each time "Proceed" is pressed after Jenkins restart. It looks like we are in the exact same workflow, we...

> I am having the same issue, any one solved it yet? Sorry to say that we "resolved" this by re-coding the entire jenkins->gitlab flow using a groovy shared lib..

Are you sure your file is accessible ? ansible-inventory is working fine with your file and I can start the cluster playbook without any issue using your hosts.yml file

I would be incline to say it's a kubespray and not etcd issue. Maybe something to add/fix in https://github.com/kubernetes-sigs/kubespray/blob/4726a110fc6f6428a1caf04168ee9ded198a833f/roles/kubespray-defaults/defaults/main.yaml#L580