Version 12 (modified by, 4 years ago) (diff)


This tutorial is outdated please refer to this version

Understanding the GENI AM API using a simple Hadoop experiment


This tutorial is based on a Hadoop in a Slice tutorial written by Paul Ruth.


Run a simple Hadoop experiment and understand how resources are described and reserved in GENI. Learn how to customize those resources and automate your experiment.

While doing this experiment, use and understand the following terms: resource specification (RSpec), sliver, and aggregate manager API (AM API)

What you will learn

At the end of this exercise, you will be able to:

  • Use and understand the following terms: Resource Specification (aka RSpec), sliver, Aggregate Manager API (AM API)
  • Customize the resources/RSpec you reserve using standard mechanisms. You will:
    • understand that an RSpec is an XML file that specifies resources which are reserved using the aggregate manager API (AM API)
    • understand that all tools use common mechanisms to describe resources (RSpecs) and to reserve resources (the AM API)
    • understand how to load and edit an RSpec using Flack
      • know how to specify the image to load on a compute node
      • know how to have software automatically loaded and run on a compute node
  • Use the Omni command line tool

Experiment Topology and Setup

Tutorial Setup

Use the Omni GENI client to reserve three VMs connected by a broadcast link.

Then arrange for automatic installation and execution of software on these nodes (running Hadoop).


Background information about pre-requisites, tools, resources, and where to get help.


  • Part I: Design/Setup
    • Step 1: Design Experiment
    • Step 2: Establish Management Environment
    • Step 3: Obtain Resources
  • Part II: Execute
    • Step 4: Configure and Initialize Services
    • Step 5: Execute Experiment
    • Step 6: Analyze and Visualize Experiment

Attachments (1)

Download all attachments as: .zip