getting-started.md

Running Galaxy

Please refer to building for building docker image first or using the newest image from dockerhub.

Copy all files in yaml directory to you kubectl client machine. And then docker push your image to Docker hub or private registry server.

Starting Galaxy daemonset

# Please update image prefix in these yamls
kubectl create -f galaxy.yaml

Update Kubelet command line args

Add --network-plugin=cni --cni-bin-dir=/opt/cni/bin/ these args to kubelet and restart it.

Starting Flannel

Galaxy uses Flannel by default to create an overlay network for Pods. Please refer to flannel to install Flannel.

Please refer to Galaxy configuration if you don't wand to install Flannel and want to change the default network.

Starting an example deployment

cat <<EOF | kubectl create -f -
apiVersion: extensions/v1beta1
kind: Deployment
metadata:
 labels:
   app: nginx
 name: nginx
spec:
 replicas: 2
 selector:
   matchLabels:
     app: nginx
 template:
   metadata:
     labels:
       app: nginx
   spec:
     containers:
     - image: nginx:1.7.9
       name: nginx
EOF

Waiting for Pods to be ready and check if you can reach pods by curl http://$(pod_ip)