Installing Pivotal HD 3.0 on single-node VM using Ambari

Preparations

First we need a VM to work with. Follow the instructions in PreparingVMforAmbari.asciidoc

We can follow steps 2.2 through 2.8 from the Pivotal Ambari Installation Guide

We can follow steps 3.2 through 3.6 from the Pivotal Ambari Installation Guide

I installed the PHD-3.0.0.0 and PHD-UTILS-1.1.0 stacks.

My Repository settings are (hawaii is the hostname for my YUM repo machine):

`redhat6`	`PHD-3.0`	`http://hawaii/PHD-3.0.0.0`
	`PHD-UTILS-1.1.0.20`	`http://hawaii/PHD-UTILS-1.1.0.20`

I configured my single-node VMs hostname as the only host.

For the SSH Private key I used the one we generated on the VM. Just run this and cut-and-paste the key:

# cat .ssh/id_dsa

I ignored the warning about my hostname not beeing a FQDN.

For services to install I picked "HDFS", "YARN + MapReduce2", "Nagios", "Ganglia" and "ZooKeeper". Everything ends up on the single node.

You need to set password and alert email for Nagios.

The Hadoop cluster should now be operational.