start + inCluster flag do not work as expected about kube-vip HOT 10 CLOSED

Your suggested workaround seems to work, tnx!

from kube-vip.

Hi, So what is happening here is that the admin.conf isn't being passed to the configuration ->

https://kube-vip.io/control-plane/#generate-the-manifest
has the steps for generating a daemonset manifest, looking at your configuration, the following should generate what you need -> curl -sL kube-vip.io/k3s | vipAddress=10.8.95.3 vipInterface=eth0 sh you'll need to edit the volumes.hostPath.path to match where there kubernetes configuration will be generated.

from kube-vip.

Thanks for the quick reply! I know the admin.conf is not getting passed, it is precisely what I want. In my cluster (being Talos) this file is not present on the nodes (and its best practise to use in-cluster client from within the cluster).

So, since there is code and a flag for using in-cluster config, can you tell me how I do that?

from kube-vip.

Ah ok! that's good to know. Then this must be a bug! the inCluster is a flag not an environment variable.. so you'll need to add it to the command.

- /kube-vip
- start
- --inCluster

off the top of my head :)

from kube-vip.

Awesomesauce.. i'll leave this open to track the issue for now :-) (any other feedback is totally welcome)

from kube-vip.

This has all been addressed now in a newer release

from kube-vip.

@jonkerj I'm working to get kube-vip working in talos but not sure how to get logs if my control-plane is unavailble. Do you have any ideas where to look?

from kube-vip.

You can inspect the containers with docker/containerd?

from kube-vip.

Okay, that's a good sign the static pod never ran. talosctl containers doesn't show it

from kube-vip.

Hi @anthr76, as of Talos 0.9 one can have Talos do vip (they actually include kube-vip), which made this approach a bit obsolte for me. You probably want to use talosctl containers -k and talosctl logs -k to debug your static pods.

from kube-vip.