Added cuda-init script for a worker and gitignore file

I have added .DS_Store files to gitignore, since MacOSX throws them all over the place.

Furthermore, I added a init-worker-with-cuda.sh script, that initializes a worker with a fresh installation of cuda, nvidia drivers and more.

I hope that this will solve a few bugs.

.gitignore

@@ -0,0 +1,2 @@
+## .DS_Store files of MacOsX
+*.DS_Store

README.md

@@ -31,8 +31,8 @@ Be aware that the following sections might be opinionated. Kubernetes is an evol
         - [Use fast setup script](#fast-track---setup-script)
         - [Manually step by step instructions](#detailed-step-by-step-instructions)
   * [How to build your GPU container](#how-to-build-your-gpu-container)
-        - [Essential parts of .yml](#essential-parts-of-yml)
-        - [Example GPU deployment](#example-gpu-deployment)
+    - [Essential parts of .yml](#essential-parts-of-yml)
+    - [Example GPU deployment](#example-gpu-deployment)
   * [Some helpful commands](#some-helpful-commands)
   * [Acknowledgements](#acknowledgements)
   * [Authors](#authors)
@@ -180,7 +180,7 @@ I will use wave-works, just because of my personal preference ;)
 ```
 kubectl apply -f https://git.io/weave-kube-1.6
 ```
-**5.II** You are ready to go, now check that pods that all pods are online to confirm that everything is working ;)
+**7** You are ready to go, now check that pods that all pods are online to confirm that everything is working ;)
 ```
 kubectl get pods --all-namespaces
 ```
@@ -280,7 +280,7 @@ export KUBECONFIG=~/.kube/admin.conf
 ```
 You are ready to use kubectl on you local client.
 
-**3.II** You can test by listing all your pods
+**4** You can test by listing all your pods
 ```
 kubectl get pods --all-namespaces
 ```

scripts/init-worker-with-cuda.sh

@@ -0,0 +1,42 @@
+#!/bin/sh
+# Following arguments are necessary:
+# 1. -> token ,  e.g. "f38242.e7f3XXXXXXXXe231e"
+# 2. -> IP:port of the master
+echo "The following token will be used: ${1}"
+echo "The master nodes IP:Port is: ${2}"
+sudo bash -c 'apt-get update && apt-get install -y apt-transport-https
+curl -s https://packages.cloud.google.com/apt/doc/apt-key.gpg | apt-key add -
+cat <<EOF >/etc/apt/sources.list.d/kubernetes.list
+deb http://apt.kubernetes.io/ kubernetes-xenial main
+EOF
+apt-get update'
+sudo apt-get install -y --allow-unauthenticated docker-engine
+sudo apt-get install -y --allow-unauthenticated kubelet kubeadm kubectl kubernetes-cni
+sudo groupadd docker
+sudo usermod -aG docker $USER
+
+# Install Cuda and Nvidia driver
+sudo apt-get install linux-headers-$(uname -r)
+sudo add-apt-repository ppa:graphics-drivers/ppa
+sudo apt-get update
+sudo apt-get install nvidia-375
+sudo apt-get install nvidia-cuda-dev nvidia-cuda-toolkit nvidia-nsight
+
+sudo systemctl enable docker && systemctl start docker
+sudo systemctl enable kubelet && systemctl start kubelet
+
+echo 'You might need to reboot / relogin to make docker work correctly'
+
+for file in /etc/systemd/system/kubelet.service.d/*-kubeadm.conf
+do
+    echo "Found ${file}"
+    FILE_NAME=$file
+done
+
+echo "Chosen ${FILE_NAME} as kubeadm.conf"
+sudo sed -i '/^ExecStart=\/usr\/bin\/kubelet/ s/$/ --feature-gates="Accelerators=true"/' ${FILE_NAME}
+
+sudo systemctl daemon-reload
+sudo systemctl restart kubelet
+
+sudo kubeadm join --token $1 $2