-
Felix from Calico exposes metrics in Prometheus format, we should be collecting them via ServiceMonitor object with Prometheus Operator.
However, during cluster bootstrapping, ServiceMonitor CRD is…
-
We should have a simple "overview" README in places like this: https://github.com/kinvolk/lokomotive/tree/master/docs/quickstarts
This README should list the contents of the directory with a short …
-
## General
We should consider adding validation for cluster configuration. At the moment we catch *some* configuration problems (mainly wrong HCL syntax), however we probably want to have more Loko…
-
### Your current KubeKey version
v1.1.1
### Describe this feature
Given my good experiences with KubeSphere I wanted to start using KubeKey to provision more bare-bones Kubernetes clusters as well.…
-
We have above variable only for worker nodes. We should add that to the controllers as well.
If that is already available on controllers, then please document such in https://github.com/kinvolk/lok…
-
In https://github.com/kinvolk/lokomotive/issues/160 we fixed a bug that happened on `lokoctl cluster apply` but not on `lokoctl cluster apply --skip-components` + `lokoctl component apply` like we run…
-
Once access to packet machines over public interface is disabled, users can still ssh into the machine using private IP of the machine and doorman vpn setup.
Docs: https://www.packet.com/developers…
-
Lokomotive uses a self-hosted control plane based on Bootkube + Helm. The sequence of events which eventually results in an operational self-hosted k8s control plane is rather complex and contains mul…
-
## Current situation
The not well known and therefore not much used systemd services for flannel and etcd require Docker and have many hacks for that to work. However, we want to eventually move Do…
-
Good day Kinvolk team.
I'm looking at lokomotive k8s and wonder if you have anything similar to [GKE Workload Identity](https://cloud.google.com/kubernetes-engine/docs/how-to/workload-identity) wh…