Large Data Virtualization instance: Tarball deployment of the Hadoop release

In the blog "Agile Management of the various releases of Hadoop", we introduced the vsphere Big Data Extensions (BDE) is to solve the enterprise deployment and management of the Hadoop release of the weapon, It makes it easy and reliable to transport the many mainstream commercial distributions of Hadoop (including the community's Apache Hadoop). This article focuses on the topics above to illustrate how to flexibly deploy and manage the Hadoop release.

Example: Deploying a GPHD1.2 release in tarball mode

First, you need to download the Gphd 1.2 release to Local. Here is the download address for Tarball and unzip the file when it is finished. You can see a set of directories:

For this example, we're going to deploy a yellow-highlighted component.

SSH to the BDE server with Serengeti account, Windows recommends using the WinSCP tool

In the file system, navigate to/opt/serengeti/www/distros/, create a new directory, named gphd/1.2.0.0/

Upload the tar files from the three folders of Hadoop, pig, and hive in the local Gphd 1.2 release to the gphd/1.2.0.0/directory, as shown in the figure: