-
I have created a vSphere-cluster in a VxRail vCenter (vSphere Client version 7.0.2.00500) I'm running clusterctl and vsphere v1.0.2. I then do $ helm template ingress-nginx ingress-nginx/ingress-nginx > nginx.yaml
$ kubectl apply -f nginx.yaml Everything starts fine, logs looks resonably ok, except regarding the loadbalancer-service... then, (waiting 1+ hour) I still get: $ kubectl get svc
NAMESPACE NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
default ingress-nginx-controller LoadBalancer 10.103.209.60 <pending> 80:31925/TCP,443:30602/TCP 5h53m
default ingress-nginx-controller-admission ClusterIP 10.108.172.48 <none> 443/TCP 5h53m I dont know if I've missed something, if something is missing from my setup, if it's a bug, or missing feature... or what it could be. The documentation regarding Tanzu is not helping me much either, since most of those pages talk about the "haproxy" (deprecated in cluster-api), which i'm not sure is even related to this issue (that seems to be regarding HA on the api-server) I have tried creating a Calico-resource: apiVersion: crd.projectcalico.org/v1
kind: BGPConfiguration
metadata:
name: default
spec:
serviceLoadBalancerIPs:
- cidr: 10.4.14.104/31 This didn't change anything, and the calico-controller dont log detecting a new LoadBalancer-service (annotated with matching IP) I've also tried using MetalLB but that didn't help either. Any help is appreciated. |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment
-
Result: |
Beta Was this translation helpful? Give feedback.
Result:
CAPV doesn't support (out-of-scope) regarding LoadBalancer.
I'm chaning this to a NodePort and dong tooling to update our external loadbalancer.