=======

Objectpools BETA

Final release name is likely to be ObjectTiering

Put all stuff in here that helps to test, deploy, automate and run the ObjectTiering feature.

Repository is maintained by [email protected] and can be updated by everyone.

Creating a set of EC2 nodes for your MapR cluster and one more node, to represent a client

For the Beta-2 testing, there is no MapR-Installer support. You have to manually provision your set of nodes.

The first step of course, is to be able to create a set of nodes on AWS, to represent your MapR Cluster and at least one client node. To do that, we are going to use Ansible and its 'ec2' module.

Install Ansible

Go to ansible.com and follow the instructions to install it for your desktop/laptop.

Using Ansible Playbooks

Playbooks are the input to Ansible. It uses a special YAML format. Playbooks contain plays and plays contain tasks. Each task uses a pre-defined Ansible-module (like yum/apt/file) to do a well defined task.

All the documentation is on ansible.com.

Launch your set of nodes.

Playbook file: launch_ec2_nodes.yml

This Playbook file uses the 'ec2' Ansible-module to launch the specified set of nodes.

Here are steps that you need to do, before you can start running this playbook: Step 1:

• Open the playbook file: launch_ec2_nodes.yml
• Change the values of the variables specified in the var section, like region, instance_type etc.
• There are additional variables, parameters like COUNT, NAME, that you will be overriding on the cmd line.
• Run the playbook like this:

$ ansible-playbook -i ./hosts ./launch_ec2_nodes.yml --extra-vars "COUNT=5 NAME=rajesh-objpool2-testing"

In the above command, note the following: 'hosts' : Ansible inventory file (required) 'launch_ec2_nodes.yml': Playbook COUNT : Param in the Playbook. Specify the number of EC2 nodes required for the MapR Cluster NAME : Param in the Playbook. Tag for your resources.

Deleting your nodes

After you are done with your testing, you need to delete them via your AWS Console.

NOTE: There is a bug in the AMI, the 2nd data disk (EBS Volume) was not set to terminate automatically. After deleting your instances in the EC2 Console UI, go to EC2 Console -> Volumes and delete the Volumes that are tagged as "Available" (it means they are not connected to any instance).