Closed ktyle closed 5 years ago
@ktyle I created a cluster now and it completed successfully, I wonder if it was a temporary error. As you are doing testing on your current deployment, you could also launch another cluster (just change the name) and see if that succeeds.
will do. In case it fails, I did do the following steps after installing kubectl on my local server that I'm running openstack on:
mkdir kubectl_secret cd kubectl_secret openstack coe cluster config k8s
But that returns "cluster in status CREATE_FAILED" so no config / certificate files get retrieved.
ok, so better remove the cluster and launch it again. We cannot easily get around that error pretending it worked.
If it fails again, try to connect as soon as possible to the master VM and look through:
kubectl get all --all-namespaces
to check if anything is failing and grab their logs.
@zonca I think you were right that this was a temporary error! I have tried the cluster creation both yesterday evening and today and they succeeded. I will go ahead and proceed further with the deployment. Feel free to close.
great, thanks!
I am following the steps as in https://zonca.github.io/2019/06/kubernetes-jupyterhub-jetstream-magnum.html to deploy a Kubernetes cluster on my Jetstream/IU allocation.
When I run bash create_cluster.sh, after about 30-40 minutes, the task fails (see the "faults" field) with an authorization error that appears tied to the Docker container.
However, the two VM's get created and I can ssh into them: openstack server list
Here is the output of create_template:
Wondering what needs to be done to resolve the authorization error.
--Kevin