EXTERNAL-IP stays in <pending> state

[dwschulze]

When I applied your .yaml to my two node cluster (one master, one worker) the external-ip stays in Here's the service and pod:

NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
hello-kubernetes LoadBalancer 10.110.156.22 80:30210/TCP 18m
$ kubectl get pods
NAME READY STATUS RESTARTS AGE
hello-kubernetes-594f6f475f-4s9nw 1/1 Running 0 18m
hello-kubernetes-594f6f475f-7zb4h 1/1 Running 0 28s
hello-kubernetes-594f6f475f-jlrwb 1/1 Running 0 28s

Browser requests to port 30210 on either host is successful, but I should still be able to get an external-ip to connect on port 8080.

Do I need to change something in the .yaml to run on my two node cluster?

[shawndwells]

Able to replicate on stock k3s install:

$ sudo kubectl get pods
NAME                                READY   STATUS    RESTARTS   AGE
svclb-hello-kubernetes-npwb2        0/1     Pending   0          48s
hello-kubernetes-594f6f475f-gpfvs   1/1     Running   0          48s
hello-kubernetes-594f6f475f-tss77   1/1     Running   0          48s
hello-kubernetes-594f6f475f-kclvr   1/1     Running   0          48s

$ sudo kubectl get service
NAME               TYPE           CLUSTER-IP     EXTERNAL-IP   PORT(S)        AGE
kubernetes         ClusterIP      10.43.0.1      <none>        443/TCP        13m
hello-kubernetes   LoadBalancer   10.43.160.18   <pending>     80:32474/TCP   90s

Am able to connect to https://:32474

[bludkiewicz]

Hey all,

I also just ran into this issue on my vanilla k3s install. From the docs:

If you try to create a Service LB that listens on port 80, the Service LB will try to find a free host in the cluster for port 80. If no host with that port is available, the LB will stay in Pending.

I switched the port for the load balancer in the yaml to 8080 and it works for me now. FYI.

[paulbouwer]

The recently added Helm Chart and Ingress options should provide options that work across more platforms.