Using a vSwitch to connect a dedicated server lets route creation fail.
Warning FailedToCreateRoute 2m36s (x8272 over 23h) route_controller (combined from similar events): Could not create route e1ced4d4-b410-4b86-9138-72c1ede017b6 10.217.4.0/24 for node starfleet-janeway after 166.14063ms: hcloud/CreateRoute: hcops/AllServersCache.ByName: hcops/AllServersCache.getCache: not found
I am running hcloud-cloud-controller-manager w/network v1.8.1. Cilium v1.8.2 is the CNI plugin, k8s 1.19.4. It would be great to just have a special label for the nodes, that should be ignored my the controller manager. Actually a couple of days ago it was still possible to create a route to the vSwitch in the web UI and with hcloud. Something changed on your end?
Using a vSwitch to connect a dedicated server lets route creation fail.
Warning FailedToCreateRoute 2m36s (x8272 over 23h) route_controller (combined from similar events): Could not create route e1ced4d4-b410-4b86-9138-72c1ede017b6 10.217.4.0/24 for node starfleet-janeway after 166.14063ms: hcloud/CreateRoute: hcops/AllServersCache.ByName: hcops/AllServersCache.getCache: not found
I am running hcloud-cloud-controller-manager w/network v1.8.1. Cilium v1.8.2 is the CNI plugin, k8s 1.19.4. It would be great to just have a special label for the nodes, that should be ignored my the controller manager. Actually a couple of days ago it was still possible to create a route to the vSwitch in the web UI and with hcloud. Something changed on your end?
Hetzner network: 10.0.0.0/24 vSwitch : 10.0.1.0/24 CNI CIDR: 10.217.0.0/16 Dedicated server: 10.0.1.2 Route: gateway 10.0.1.2 destination 10.217.4.0/24
Please don't close the issue, because not allowing it results in taints for the node.
node.kubernetes.io/unreachable:NoExecute node.kubernetes.io/network-unavailable:NoSchedule node.kubernetes.io/unreachable:NoSchedule
The cloud controller manager must ignore the node, so that I can manually create the routes.