install_spark.sh

#!/bin/bash

# Create namespace "spark" where will run Jobs
kubectl create namespace spark --dry-run=client -o yaml | kubectl apply -f -

# Create service account "spark" to spark-operator allow submit jobs
kubectl create serviceaccount spark -n spark --dry-run=client -o yaml | kubectl apply -f -
kubectl create clusterrolebinding spark-role --clusterrole=edit --serviceaccount=spark:spark --namespace=spark --dry-run=client -o yaml | kubectl apply -f -

# Install spark operator
helm upgrade --install spark-operator spark-operator/spark-operator --debug \
    --namespace spark-operator \
    --create-namespace \
    -f spark/helm-release/values.yaml

kubectl create configmap -n spark-operator spark-config --from-file=./spark/docker/spark-custom/conf/spark-defaults.conf --dry-run=client -o yaml | kubectl apply -f -