Understanding how the different components of a vFabric Data Director for Hadoop environment are related helps you manage and troubleshoot your system.

You deploy the vFabric Data Director for Hadoop OVA in a vSphere environment. After deployment, a management server virtual machine and a template virtual machine are available.

The management server orchestrates deployment and runs a service that you can access from the Data Director for Hadoop Web Console.

The template virtual machine is cloned each time an additional Hadoop node is created and added to the cluster.

When you create a Hadoop cluster, the management server performs these steps.

1

Searches for ESXi hosts with sufficient resources.

2

For each Hadoop virtual machine (node), selects an ESXi host on which to place the virtual machine.

3

Sends a request to the vCenter Server system to clone the template virtual machine, configure the clone, and place the clone on the specified ESXi host.

4

The agent that is part of each virtual machine that is cloned from the template virtual machine configures the network and other operating system parameters.

5

When the user adds a Hadoop distribution to vFabric Data Director for Hadoop, the software is installed only on the management server.

6

When you create a cluster with a specific Hadoop distribution, the agent on each virtual machine copies the software components required for that node type from the management server to the virtual machine.

7

Finally, the agent configures the necessary Hadoop parameters.

The vFabric Data Director for Hadoop OVA includes a management server. The management server has functionality for managing Hadoop. You do not use other Data Director functionality, such as managing resource bundles or managing environment, when you work with vFabric Data Director for Hadoop. Instead, you access the Data Director for Hadoop Web Console, either directly or from the Data Director Web console. See Access the Data Director for Hadoop Web Console from Data Director. You can also download the Serengeti CLI Console from the Data Director for Hadoop login page. See Install the Data Director for Hadoop CLI Console