Deploying Scylla on GKE¶

Configure environment variables¶

First of all, we export all the configuration options as environment variables. Edit according to your own environment.

GCP_USER=$( gcloud config list account --format "value(core.account)" )
GCP_PROJECT=$( gcloud config list project --format "value(core.project)" )
GCP_REGION=us-west1
GCP_ZONE=us-west1-b
CLUSTER_NAME=scylla-demo
CLUSTER_VERSION=$( gcloud container get-server-config --zone ${GCP_ZONE} --format "value(validMasterVersions[0])" )

Creating a GKE cluster¶

First we need to change kubelet CPU Manager policy to static by providing a config file. Create file called systemconfig.yaml with the following content:

kubeletConfig:
  cpuManagerPolicy: static

Then we’ll create a GKE cluster with the following:

1. A NodePool of 4 n1-standard-32 Nodes, where the Scylla Pods will be deployed. Each of these Nodes has 8 local SSDs attached, which will later be combined into a RAID0 array. It is important to disable autoupgrade and autorepair, since they increase the frequency of loss of data on local SSDs.

   gcloud container --project "${GCP_PROJECT}" \
   clusters create "${CLUSTER_NAME}" --username "admin" \
   --zone "${GCP_ZONE}" \
   --cluster-version "${CLUSTER_VERSION}" \
   --node-version "${CLUSTER_VERSION}" \
   --machine-type "n1-standard-32" \
   --num-nodes "4" \
   --disk-type "pd-ssd" --disk-size "20" \
   --local-ssd-count "8" \
   --node-taints role=scylla-clusters:NoSchedule \
   --image-type "UBUNTU_CONTAINERD" \
   --system-config-from-file=systemconfig.yaml \
   --enable-stackdriver-kubernetes \
   --no-enable-autoupgrade \
   --no-enable-autorepair
   

2. A NodePool of 2 n1-standard-32 Nodes to deploy cassandra-stress later on.

   gcloud container --project "${GCP_PROJECT}" \
   node-pools create "cassandra-stress-pool" \
   --cluster "${CLUSTER_NAME}" \
   --zone "${GCP_ZONE}" \
   --node-version "${CLUSTER_VERSION}" \
   --machine-type "n1-standard-32" \
   --num-nodes "2" \
   --disk-type "pd-ssd" --disk-size "20" \
   --node-taints role=cassandra-stress:NoSchedule \
   --image-type "UBUNTU_CONTAINERD" \
   --no-enable-autoupgrade \
   --no-enable-autorepair
   

3. A NodePool of 1 n1-standard-8 Node, where the operator and the monitoring stack will be deployed.

   gcloud container --project "${GCP_PROJECT}" \
   node-pools create "operator-pool" \
   --cluster "${CLUSTER_NAME}" \
   --zone "${GCP_ZONE}" \
   --node-version "${CLUSTER_VERSION}" \
   --machine-type "n1-standard-8" \
   --num-nodes "1" \
   --disk-type "pd-ssd" --disk-size "20" \
   --image-type "UBUNTU_CONTAINERD" \
   --no-enable-autoupgrade \
   --no-enable-autorepair
   

Setting Yourself as cluster-admin¶

(By default GKE doesn’t give you the necessary RBAC permissions)

Get the credentials for your new cluster

gcloud container clusters get-credentials "${CLUSTER_NAME}" --zone="${GCP_ZONE}"

Create a ClusterRoleBinding for your user. In order for this to work you need to have at least permission container.clusterRoleBindings.create. The easiest way to obtain this permission is to enable the Kubernetes Engine Admin role for your user in the GCP IAM web interface.

kubectl create clusterrolebinding cluster-admin-binding --clusterrole cluster-admin --user "${GCP_USER}"

Installing Helm¶

If you don’t have Helm installed, Go to the helm docs to get the binary for your distro.

Install RAID DaemonSet¶

To combine the local disks together in RAID0 arrays, we deploy a DaemonSet to do the work for us.

kubectl apply -f examples/gke/raid-daemonset.yaml

Install the local provisioner¶

After combining the local SSDs into RAID0 arrays, we deploy the local volume provisioner, which will discover their mount points and make them available as PersistentVolumes.

helm install local-provisioner examples/common/provisioner