OpenStack Installation, Configuration & Administration

OpenStack is an open-source cloud computing platform that organizations use to set up and manage their own cloud infrastructure similar to Amazon Web Services (AWS), Microsoft Azure, or Google Cloud.

By using OpenStack, you can manage IT resources including compute, storage, networking, and more. It helps manage VM templates, identities, and deploy and manage complex applications. According to Mordor Intelligence, the OpenStack services market will reach $155.36 billion in 2031, with a CAGR of 31.45%.

We’ll walk you through OpenStack installation, deployment, and administration in this blog. Let’s get started.

Table of Contents

• OpenStack Installation Methods
• Sunbeam - Single-node Installation
• Compute Node and Neutron Configuration Methods
• OpenStack Deployments
• Creating a Sandbox Compute Server with VirtualBox and Vagrant
• How to Configure Ubuntu Cloud Archive
• How to Install Compute Controller Services
• FAQs
• Conclusion

OpenStack Installation Methods

Below are the methods for installing OpenStack:

Sunbeam - Single-node Installation

Let’s begin by learning how to install OpenStack using Sunbeam in a single-node environment.

Sunbeam is the easiest way to set up a quick OpenStack cloud on Ubuntu. Canonical's documentation uses the OpenStack snap and the sunbeam command. A basic single-node deployment assigns the control, compute, and storage roles to a single machine.

Prerequisites

You need to use a fresh Ubuntu Server 24.04 LTS system. You also need to allocate sufficient CPU, RAM, and disk space. It’s best to have a separate unused disk for storage.

Install the OpenStack snap

Prepare the node

Bootstrap the cloud

Configure OpenStack

Launch a test VM

You can use the above steps to install OpenStack using the sunbeam method.

Compute Node and Neutron Configuration Methods

We’ll learn about compute node and neutron configuration methods here.

1. Compute Node Configuration

A Nova compute node hosts and manages the virtual machines (instances) that run in an OpenStack cloud. The nova-compute service interacts with the control plane via RabbitMQ and the Nova API, as well as the conductor service. The underlying hypervisor, KVM/QEMU with libvirt, creates and manages the instances.

For a typical Ubuntu-based deployment, the setup includes the following:

2. Network (Neutron) Configuration

OpenStack Neutron provides networking as a service. It manages networks, subnets, ports, routers, security groups, floating IPs, and connections between instances and external networks.

In a typical OpenStack setup, networking tasks are divided between controller and/or network nodes and compute nodes. The specific setup varies based on the selected networking architecture, such as OVN with Open vSwitch, Open vSwitch with agents, or Linux Bridge.

OpenStack Deployments

Let’s discuss the various OpenStack deployment methods below.

1. OpenStack Deployment with Ansible

OpenStack deployment with Ansible automates the installation, configuration, and management of a private OpenStack cloud environment.

Ansible uses playbooks and inventory files to configure controller, compute, network, and storage nodes consistently. It can automate the deployment of major OpenStack services, including Keystone, Nova, Neutron, Glance, Cinder, and Horizon.

A common method is Kolla-Ansible, which uses Ansible to deploy OpenStack services in containers. The process includes the following:

Ansible templates generate service configuration files, while variables help administrators customize networking, storage, authentication, and other deployment settings. After deployment, administrators can verify OpenStack services, create projects and users, set up networks, upload images, and launch virtual machines.

This deployment method reduces manual setup, improves consistency, and makes OpenStack environments easier to reproduce and manage.

2. Deployment with Kolla-Ansible

OpenStack deployment with Kolla-Ansible provides an easy way to deploy and manage OpenStack services using Ansible and containers.

Kolla-Ansible automates the setup and deployment of OpenStack components across controller, compute, network, and storage nodes. It employs Docker to run OpenStack services in separate containers.

Let’s go through the highlights of this deployment.

• The deployment starts with preparing the host operating systems, configuring networking and SSH access, and installing Ansible and Kolla-Ansible.
• You can create an inventory file to define the OpenStack nodes and to set deployment variables such as network interfaces, container registries, storage options, and service settings.
• You can use the Kolla-Ansible commands to set up the servers, perform pre-deployment checks, deploy the containers, and configure OpenStack.
• You can deploy key services such as Keystone, Nova, Neutron, Glance, Cinder, and Horizon using this automated process.
• After deployment, you can use the OpenStack CLI or Horizon to create projects, users, networks, images, and virtual machines.

Kolla-Ansible also simplifies upgrades, configuration changes, and operational management.

Creating a Sandbox Compute Server with VirtualBox and Vagrant - OpenStack

VirtualBox allows users to spin up virtual machines and networks without affecting other workloads. It is available for free on Windows, macOS, and Linux. By creating a sandbox environment with VirtualBox and Vagrant, we can learn about and experiment with OpenStack.

1. Creating a Sandbox Compute Server

Creating a sandbox server to run OpenStack Compute services is straightforward with VirtualBox and Vagrant. It allows you to experiment with OpenStack services.

VirtualBox lets us spin up virtual machines and networks without affecting your working environment. It is available for free for Windows, macOS, and Linux.

On the other hand, Vagrant allows you to automate tasks, so you can spend less time creating your test environments and more time using OpenStack. Vagrant can be installed using Ubuntu’s package manager.

Importantly, you must ensure that the computer you use to run your test environment has sufficient processing power. It should include hardware virtualization support for modern AMD and Intel processors, as well as at least 8 GB of RAM.

2. Getting Started

To create your sandbox server for running OpenStack Compute in VirtualBox, you can use Vagrant to provision a second virtual machine that hosts OpenStack Compute instances.

This virtual machine, or the OpenStack Compute node, can be configured with at least 3 GB of RAM, 2 CPUs, 20 GB of disk space, and 3 network interfaces.

The three interfaces are:

• A NAT interface allows the virtual machine to connect to the network outside of VirtualBox to download packages
• A public interface for the OpenStack Compute host
• A private network interface that OpenStack Compute uses for internal communication between different OpenStack Compute hosts.

3. The Procedure to Create a Sandbox Server

Below is the procedure to create a sandbox server:

How to Configure Ubuntu Cloud Archive (UCA) in OpenStack

Ubuntu Cloud Archive (UCA) provides OpenStack packages for supported Ubuntu releases. For current deployments, Ubuntu 26.04 LTS is the latest Ubuntu LTS release.

Canonical's current OpenStack deployment guidance uses newer tools such as Sunbeam. Additionally, Canonical's OpenStack reference architecture indicates that Ubuntu 26.04 LTS is supported with OpenStack 2026.1 LTS.

The current deployment workflow is as follows:

Install Ubuntu Server 26.04 LTS on the OpenStack host.

Then, update the operating system using the following commands.

Verify the Ubuntu release using the command below

Enable the required networking, hostname resolution, and NTP/time synchronization, SSH access, and storage configurations.

Install the OpenStack snap for modern Canonical OpenStack deployment:

Implement the deployment and configuration steps relevant to the chosen OpenStack architecture.

For Ubuntu 24.04 LTS (Noble Numbat), the Ubuntu Cloud Archive supports Epoxy, which is linked to the OpenStack 2025.1 release. Ubuntu added “cloud-archive:epoxy” support to software-properties and backported the support to Noble.

For Ubuntu 22.04 LTS (Jammy Jellyfish), OpenStack 2024.1 (Caracal) is available in the Ubuntu Cloud Archive. Canonical's current Cloud Archive documentation specifically lists Caracal for Ubuntu 22.04.

How to Install Compute Controller Services – OpenStack

Compute is a collection of services that enable you to launch virtual machine instances. Most services run on a controller node, and the service responsible for launching virtual machines runs on a dedicated compute node.

Let’s see how to install and configure these services on the controller node.

OpenStack Compute (Nova) is the compute service of OpenStack. It is the component that allows you to run multiple virtual machine instances on any number of hosts running the OpenStack Compute service. It enables a highly scalable and redundant cloud environment.

OpenStack Compute is a major part of an Infrastructure-as-a-Service (IaaS) system. It interacts with OpenStack Identity for authentication, OpenStack Image Service for disk images and server images, and the OpenStack dashboard for the user and administrative interface.

A compute node handles connectivity and security groups for an instance.

The configuration file, /etc/nova/nova, specifies how to run each OpenStack Compute service and which services to connect to when presenting OpenStack to end users.

This file will be replicated among your nodes as your environment grows.

OpenStack Compute powers some compute clouds, such as Rackspace Open Cloud. You must use the default Nova Networking VLAN Manager in multi-host mode, not Neutron Software-defined networking.

Separating controller services from the compute nodes allows you to scale your OpenStack environment and resources horizontally in the controller and compute services.

You need to install additional packages on your controller node, including Keystone OpenStack Identity Services and the OpenStack Image Service.

Before you create a server to run OpenStack compute services for your instances, you need to install some services on the controller node where the OpenStack Identity and Image services are running.

The services include:

• nova-scheduler:  The scheduler selects a suitable host to fulfill the request to run the instance.
• nova-api: The service that handles API requests to OpenStack to operate the services within it; for example, you make a call to this service to start up a new Nova instance.
• nova- conductor: A new service introduced in Grizzly to remove direct database calls by the Compute service.
• nova-objectstore: It is a file storage service
• nova-common: Common Python libraries required for the OpenStack environment.
• nova-cert: The Nova certificate management service you can use to authenticate to Nova.
• ntp: Network Time Protocol is essential in a multi-node environment. It ensures nodes have synchronized time.
• dnsmasq: A DNS forwarder and DHCP service that allocates IP addresses to your instances.

Getting Started with OpenStack Controller Node

Ensure that you are logged in to the OpenStack controller node. 

• Installation of OpenStack

Install OpenStack on Ubuntu 26.04 LTS.

• Installation of OpenStack Compute Controller

Installing OpenStack Compute Controller packages from the Ubuntu Cloud Archive repository is a simple way to get the latest OpenStack on your Ubuntu server.

Frequently Asked Questions

1. Is it easy to learn OpenStack?

Yes, it is easy to learn OpenStack. If you have a basic understanding of software installation and configuration, learning the OpenStack process will be easier.

2. What is the current OpenStack release?

The latest version of OpenStack is 2026.1 Gazpacho as of August 19, 2026.

3. How does OpenStack compare to AWS for private cloud deployments?

You can deploy OpenStack in your own data centre, whereas AWS is mainly a provider of public cloud services. Although AWS offers services such as AWS Outposts that enable the setup of environments similar to private clouds, this is not the same as operating a native OpenStack private cloud.

4. What replaced VirtualBox/Vagrant for OpenStack sandboxes?

KVM/libvirt, together with Kolla-Ansible, is a good option for setting up a realistic OpenStack sandbox, while DevStack is the better choice for OpenStack development.

Conclusion

This tutorial has given you a clear understanding of the OpenStack platform and its installation. You have also learned in detail about the Ubuntu Cloud Archive and compute controller services.

If you want to learn more about OpenStack, you can sign up for an OpenStack course with MindMajix. By the end of the training, you will gain the skills to manage OpenStack in complex environments effectively.