Eric Greer

Makes sense. We should definitely include the security settings in all of our `khcheck` specs deployed with the helm chart. As a workaround, you could use `helm template` and then...

Hrmm... did you get any errors when you applied your helm chart? This RBAC seems to indicate the `get` permission for `nodes`: https://github.com/kuberhealthy/kuberhealthy/blob/master/deploy/helm/kuberhealthy/templates/clusterrole.yaml#L59 Can you run `kubectl get clusterrole kuberhealthy`...

I think namespace selectors would be a good addition to our Kuberhealthy checks. Because of the highly customizable nature of a Kuberhealthy check, this would need implemented on every check....

Awesome! Hopefully that is approved and such. We need a certain number of companies adopting Kuberhealthy to promote to the next level within the CNCF.

Thinking about this a little more... There are some more things that would need considered. Each `khcheck` resource is considered one check that kuberhealthy manages. This results in one registered...

Thanks for the carefully thought through proposal here. My initial impression is that this would be complicated due to Kuberhealthy not having a database of any kind. The metrics would...

I think this is do-able. Right now, Kuberhealthy does a reload on all checks every time anything changes to ensure that all state is correct and to keep the code...

This is awesome! I will code review this tomorrow. Thanks for providing such great benchmark data. Sounds like this is important in order to scale to large numbers of metrics...

This looks pretty wild. Can you check the logs from the Kuberhealthy pods around the same time to see if they are reporting why these status reports are being denied?...

Check the events on the kuberhealthy pods and see if they are restarting... You can also run `kubectl get ep kuberhealthy` to check the endpoints to see if the pods...