Cluster managers
Cluster managers are used to deploy Spark applications in cluster mode. Spark can be configured to run various cluster managers. Spark distribution provides an inbuilt cluster manager known as Spark standalone. Apart from that Spark can run on top of other popular cluster managers in the big data world such as YARN and Mesos. In this section, we will discuss how to deploy Spark applications with Spark standalone and YARN.
Spark standalone
Spark standalone manager is available in the Spark distribution. It helps to deploy Spark applications in cluster mode in a very efficient and convenient way.
Spark standalone manager follows the master-slave architecture. It consists of a Spark master and multiple worker nodes where worker nodes are the slave nodes for Spark master node. Similar to other master-slave frameworks, Spark master works a scheduler for the submitted Spark applications. It schedules the applications on worker nodes and the processes that executed the application...