Before continuing, be sure you have completed stage 1 - stage 3.

At this point your Kubernetes cluster is installed. We will run a few scripts to ensure that you can now start doing work on your cluster.

1. Verify up

Log into kube01 and verify you are up and running:

ssh kube01
kubectl get componentstatuses

You should see etcd healthy.

kubectl get nodes

You should see three nodes up

NAME      STATUS         AGE
kube01    Ready,master   26m
kube02    Ready          23m
kube03    Ready          23m

2. Kubectl

You may decide you want to access your cluster remotely. This requires the kubectl command tool to be installed on the machine you want to access from as well as creating a file on your build server (or MacOS) and set kubectl to use it.

2.1 Get kubectl

Instructions are posted here. We recommend version 1.5.4 as this is the version that works with contiv.

curl -LO https://storage.googleapis.com/kubernetes-release/release/v1.5.4/bin/linux/amd64/kubectl
chmod 755 kubectl
sudo mv kubectl /usr/local/bin

2.2 Config Kubectl

You create a file called:

~/.kubectl/config

Most of the contents of the file can be retrieved from kube01 in /etc/kubernetes/admin.conf To make it simple just run the following:

mkdir -p ~/.kube
scp kube01:/etc/kubernetes/admin.conf ~/.kube/config

After doing this, kubectl should work right away:

kubectl get componentstatus

2.1 troubleshooting

If you are using proxies, make sure that you have the kubernetes master node configured in your ~/.bash_profile as no_proxy, example:

export no_proxy=10.61.124.100,10.61.124.170

In the above example, this is set with the build server IP address and the kube01 IP address.