Recipes are not yet available on the Cloudbreak UI, thus if you’d like to try this Spark plugin, we suggest to use the Cloudbreak Shell.
Clone the project from GitHub.
git clone https://github.com/sequenceiq/consul-plugins-spark.git
From Cloudbreak shell (after the usuall steps - use hint):
recipe add --file /consul-plugins-spark/spark-recipe.json
Alternatively you can use the --url parameter to create the Spark recipe.
recipe add --url https://raw.githubusercontent.com/sequenceiq/consul-plugins-spark/master/spark-recipe.json
For using your recipe you have to select it. (you can check your recipes with recipe list command)
recipe select --id RECIPE_ID
In case of your blueprint and credential have already been selected and the instance groups are configured (instancegroup configure), create your stack with create stack command. If your stack is ready, you are able to create your cluster:
cluster create --description "Cloudbreak provisoned cluster with Apache Spark"
Note that: this recipe is for using Apache Spark with YARN (yarn-cluster or yarn-client mode)