Shutting down Spark clusters may not be as critical when working in a local environment; however, it will prove costly when Spark is deployed in a cloud environment where you are charged for compute power.
