Running on a native Google (GCP) Kubernetes Clusters - phnmnl/phenomenal-h2020 GitHub Wiki

To run on a GCP native Kubernetes:

• Create k8s cluster from Google console or through CLI
• Go to the connect part on GCP console, it will provide a gcloud CLI command to setup kubectl to connect to the clusters
• Create following cluster-binding-role for helm (kubectl create -f) <fileWithFollowingContent>:

apiVersion: v1
kind: ServiceAccount
metadata:
  name: tiller
  namespace: kube-system
---
kind: ClusterRoleBinding
apiVersion: rbac.authorization.k8s.io/v1beta1
metadata:
  name: tiller-clusterrolebinding
subjects:
- kind: ServiceAccount
  name: tiller
  namespace: kube-system
roleRef:
  kind: ClusterRole
  name: cluster-admin
  apiGroup: ""

• helm init on a machine where you have previously added galaxy-helm-repo helm repository.
• You might need:

kubectl patch deploy --namespace kube-system tiller-deploy -p '{"spec":{"template":{"spec":{"serviceAccount":"tiller"}}}}'

• Create single node file system on GCP: https://console.cloud.google.com/launcher/details/click-to-deploy-images/singlefs?project=phenomenal-gcp-testing&folder&organizationId checking that NFS serving is activated.
  • More documentation here: https://cloud.google.com/launcher/docs/single-node-fileserver?hl=en_GB&_ga=2.9020275.-756831591.1515154233
  • Once deployed, add the no-root-squash preference to /etc/exports (you will need to ssh into the machine through google console or cli, and sudo su).
• Create PV pointing to that NFS created:

apiVersion: v1
kind: PersistentVolume
metadata:
  name: pv-nfs
spec:
  storageClassName: standard
  capacity:
    storage: 20Gi
    # volumeMode: Filesystem
  accessModes:
    - ReadWriteMany
  persistentVolumeReclaimPolicy: Recycle
  nfs:
    path: /data
    server: singlefs-1-vm

• Run our helm install process.