The following will help with understanding what is happening on the nodes:

$ cat bin/node-resources.sh 
#!/bin/bash
set -euo pipefail

echo -e "Iterating...\n"

nodes=$(kubectl get node --no-headers -o custom-columns=NAME:.metadata.name)

for node in $nodes; do
  echo "Node: $node"
  kubectl describe node "$node" | sed '1,/Non-terminated Pods/d'
  echo
done

From https://github.com/kubernetes/kubernetes/issues/17512

You have to deploy heapster in your cluster to collect metrics ;)
by default its not activated on digitalocean k8s-cluster.

There is a helm chart for heapster: https://github.com/helm/charts/tree/master/stable/heapster

• Install metrics server chart

• Update the command metrics server deployment and add the following two options to the pod command :

  - --kubelet-insecure-tls
  - --kubelet-preferred-address-types=InternalIP
  

Command will be look like this :

   command:
    - /metrics-server
    - --kubelet-insecure-tls
    - --kubelet-preferred-address-types=InternalIP

Wait two or three minutes, then the “top” command should work.