The CloudLab Manual

2025-06-03 (19a48c0)

contents ← prev up next →

The CloudLab Manual🔗

The CloudLab Team

This document is also available as a single page

Acceptable Use Policy

CloudLab is a "meta-cloud"—that is, it is not a cloud itself; rather, it is a facility for building clouds. It provides bare-metal access and control over a substantial set of computing, storage, and networking resources; on top of this platform, users can install standard cloud software stacks, modify them, or create entirely new ones.

The current CloudLab deployment consists of more than 33,000 cores in over 1800 allocatable nodes distributed across three sites at the University of Wisconsin, Clemson University, and the University of Utah. CloudLab interoperates with existing testbeds including GENI and Emulab, to take advantage of hardware at dozens of sites around the world.

The control software for CloudLab is open source, and is built on the foundation established for Emulab, GENI, and Apt. Pointers to the details of this control system can be found on CloudLab’s technology page.

1 Getting Started

2 CloudLab Users

2.1 Register for an Account

2.1.1 Join an existing project

2.1.2 Create a new project

2.1.3 Setting up SSH access

2.1.4 Setting up X11

3 CloudLab and Repeatable Research

3.1 CloudLab For Artifact Evaluation

3.1.1 For Authors

3.1.2 For AEC Chairs

3.1.3 For AEC Members

4 Creating Profiles

4.1 Creating a profile from an existing one

4.1.1 Copying a Profile

4.1.2 Creating the Profile

4.1.3 Updating a profile

4.2 Creating a profile with a GUI

4.3 Capturing Disk Images for Profiles

4.4 Repository-Based Profiles

4.4.1 Updating Repository-Based Profiles

4.4.2 Branches and Tags in Repository-Based Profiles

4.5 Creating a profile from scratch

4.6 Sharing Profiles

4.7 Versioned Profiles

5 Basic Concepts

5.1.1 On-demand Profiles

5.1.2 Persistent Profiles

5.2 Experiments

5.2.1 Extending Experiments

5.4 Physical Resources

5.5 Virtual Machines

6 CloudLab for Classes

6.1 Creating Your Project on CloudLab

6.2 Having Students Join Your Project

6.3 Project-Based Classes

6.4 Assignment-Based Classes

7 Resource Reservations

7.1 What Reservations Guarantee

7.2 How Reservations May Affect You

7.3 Making a Reservation

7.4 Using a Reservation

7.5 Who Shares Access to Reservations

8 Describing a profile with python and geni-lib

8.1 A single XEN VM node

8.2 A single physical host

8.3 Two XenVM nodes with a link between them

8.4 Two ARM64 servers in a LAN

8.5 A VM with a custom size

8.6 Set a specific IP address on each node

8.7 Specify an operating system and set install and execute scripts

8.8 Profiles with user-specified parameters

8.9 Add storage to a node

8.10 Debugging geni-lib profile scripts

9 Virtual Machines

9.1.1 Controlling CPU and Memory

9.1.2 Controlling Disk Space

9.1.3 Setting HVM Mode

9.1.4 Dedicated and Shared VMs

10 Storage Mechanisms

10.1 Overview of Storage Mechanisms

10.2 Node-Local Storage

10.2.1 Specifying Storage in a Profile – Local Datasets

10.2.2 Allocating Storage in a Running Experiment

10.2.3 Persisting Local Data

10.3 Image-backed Datasets

10.4 Remote Datasets

10.5 NFS Shared Filesystems

10.6 Storage Type Summary (TL;DR)

10.7 Example Storage Profiles

10.7.1 Creating a Node-local Dataset

10.7.2 Creating an Image-backed Dataset from a Node-local Dataset

10.7.3 Using and Updating an Image-backed Dataset

10.7.4 Creating a Remote Dataset

10.7.5 Using a Remote Dataset on a Single Node

10.7.6 Using a Remote Dataset on Multiple Nodes via a Shared Filesystem

10.7.7 Using a Remote Dataset on Multiple Nodes via Clones

11 The Control Network

11.2 Ensuring your experiment nodes use the correct interfaces

11.2.1 Common causes of control network misuse

11.2.2 Empirically verifying interface use

11.3 What to do if you get a Control Network Violation email

11.4 Why is it so important to avoid the control network?

11.5 When can I use the control network?

12 Advanced Topics

12.1 Disk Images

12.3 Public IP Access

12.3.1 Dynamic Public IP Addresses

12.5 Introspection

12.5.1 Client ID

12.5.2 Control MAC

12.5.3 Manifest

12.5.4 Private key

12.5.5 Profile parameters

12.6 User-controlled switches and layer-1 topologies

12.7 Portal API

13.1 CloudLab Utah

13.2 CloudLab Wisconsin

13.3 CloudLab Clemson

13.4 Apt Cluster

14 CloudLab Kubernetes Tutorial

14.1 Objectives

14.2 Prerequisites

16.4 Logging In

14.4 Building Your Own Kubernetes Cluter

14.5 Exploring Your Experiment

14.5.1 Experiment Status

14.5.2 Profile Instructions

14.5.3 Topology View

14.5.4 List View

14.5.5 Manifest View

14.5.6 Graphs View

14.5.8 Web-based Shell

14.5.9 Serial Console

14.6 Exploring Kubernetes

14.6.1 Kubernetes Dashboard

14.6.2 Kubernetes Dashboard

14.7 Terminating the Experiment

14.8 Taking Next Steps

15 CloudLab OpenStack Tutorial

15.1 Objectives

15.2 Prerequisites

16.4 Logging In

15.4 Building Your Own OpenStack Cloud

15.5 Exploring Your Experiment

15.5.1 Experiment Status

15.5.2 Profile Instructions

15.5.3 Topology View

15.5.4 List View

15.5.5 Manifest View

15.5.6 Graphs View

15.5.8 Web-based Shell

15.5.9 Serial Console

15.6 Bringing up Instances in OpenStack

15.7 Administering OpenStack

15.7.1 Log Into The Control Nodes

15.7.2 Reboot the Compute Node

15.8 Terminating the Experiment

15.9 Taking Next Steps

16 CloudLab Chef Tutorial

16.1 Objectives

16.2 Motivation

16.3 Prerequisites

16.4 Logging In

16.5 Launching Chef Experiments

16.6 Exploring Your Experiment

16.6.1 Experiment Status

16.6.2 Profile Instructions

16.6.3 Topology View

16.6.4 List View

16.6.5 Manifest View

16.7 Brief Introduction to Chef

16.8 Logging in to the Chef Web Console

16.8.1 Web-based Shell

16.8.2 Chef Web Console

16.9 Configuring NFS

16.9.1 Exploring The Structure

16.10 Apache Web Server and ApacheBench Benchmarking tool

16.10.1 Understanding the Internals

16.11 Final Remarks about Chef on CloudLab

16.12 Terminating Your Experiment

16.13 Future Steps

17 Citing CloudLab

18 Getting Help

contents ← prev up next →