Can't connect to application using external IP about kubernetes-engine-samples HOT 4 CLOSED

@ryanmcginnis I cannot reproduce this on my cluster. Here are some things that might've gone wrong:

1. Are you adding :8080 port number to the external IP address?
2. You might have hit a quota in your GCP project. Can you send the output of kubectl describe service hello-web, this should show if the load balancer is created correctly or not.

from kubernetes-engine-samples.

I am specifying the port. Here is the output of kuebctl describe:

Name:                     hello-web
Namespace:                default
Labels:                   run=hello-web
Annotations:              <none>
Selector:                 run=hello-web
Type:                     LoadBalancer
IP:                       10.63.242.155
LoadBalancer Ingress:     104.154.217.236
Port:                     <unset>  8080/TCP
TargetPort:               8080/TCP
NodePort:                 <unset>  30237/TCP
Endpoints:                <none>
Session Affinity:         None
External Traffic Policy:  Cluster
Events:
  Type    Reason                Age   From                Message
  ----    ------                ----  ----                -------
  Normal  CreatingLoadBalancer  2m    service-controller  Creating load balancer
  Normal  CreatedLoadBalancer   1m    service-controller  Created load balancer

I think you're probably right that I've hit a quota, let me take another look at that.

from kubernetes-engine-samples.

I found the problem and I think it's interesting behavior:

Events:
  Type     Reason            Age                 From               Message
  ----     ------            ----                ----               -------
  Warning  FailedScheduling  4s (x1378 over 6h)  default-scheduler  No nodes are available that match all of the following predicates:: PodToleratesNodeTaints (1).

I had a taint for this cluster's nodes. I would have expected to be warned in the CLI or to see in the UI that the Pods weren't being scheduled because of this. But it's fair that most users would probably remember setting a taint.

Thanks for your review!

from kubernetes-engine-samples.

would have expected to be warned in the CLI or to see in the UI that the Pods weren't being scheduled because of this.

Scheduling in Kubernetes happens asynchronous of Pod creation. In this case, we actually create more indirection by deploying pods using Deployment controller. I suspect this is by design.

from kubernetes-engine-samples.