Updates name of node pool for kube-ip (#973)

We had a disk pressure issue on the static-public-ip nodes, which caused
script-exporter to be evicted. It is not possible to resize the disks of
an existing node pool. The only way to achieve this is to create a new
node pool with the desired configurations and then migrate your
workloads to this new node pool. I created a new node pool in mlab-oti
named static-external-ip and migrated the workloads to this new node
pool. This change configures kube-ip to be sure that those nodes get
static IPv4 addresses. Meanwhile I have manually updated the kube-ip
deployment in production with the change in this commit.

Author: nkinkade

k8s/prometheus-federation/deployments/kubeip.yml

@@ -32,7 +32,7 @@ spec:
             value: "prometheus-federation"
             # Nodepool to which KubeIP assigns addresses
           - name: "KUBEIP_NODEPOOL"
-            value: "static-public-ip"
+            value: "static-external-ip"
             # GCP Application Default Credentials (ADC)
             # See: https://cloud.google.com/docs/authentication/production
           - name: "GOOGLE_APPLICATION_CREDENTIALS"