English 
Français 
Find out how it works OpenStacka platform for cloud computing flexible and scalable for building and managing applications and services.
OpenStack is a highly scalable and flexible open source cloud platform that provides a cloud computing infrastructure on which you can build and manage a variety of applications and services. As a user, you benefit from access to virtual resources such as virtual machines, storage and networks, which are provisioned automatically and orchestrated by OpenStack. Thanks to its many interconnected components, such as Nova, Neutron and Cinder, OpenStack enables efficient resource management in a cloud computing environment, offering a complete, high-performance solution to meet your needs.
What is OpenStack?
A brief introduction to OpenStack
OpenStack is an open-source cloud computing platform for creating and managing private, public or hybrid cloud environments. It offers infrastructure as a service (IaaS) for provisioning and managing on-demand computing resources, such as virtual machines, storage and networks.
OpenStack's main objective is to provide a flexible, scalable and resilient cloud computing solution, giving businesses total autonomy in managing their IT resources. Thanks to its modular architecture, OpenStack enables the platform to be adapted to the specific needs of each organization, while guaranteeing a high level of interoperability between the various components.
Origins and evolution of OpenStack
OpenStack was created in 2010 by a partnership between NASA and Rackspace, a web hosting company. The two organizations decided to collaborate to develop an open-source cloud computing solution that could compete with the proprietary offerings already on the market. Their aim was to create a platform that was free, scalable and totally vendor-independent.
Since its creation, OpenStack has experienced significant growth and become one of the most popular open-source projects in cloud computing. Today, it counts thousands of contributors worldwide and is supported by a large community of users and developers. Over the years, many companies and organizations have adopted OpenStack as their cloud computing solution, contributing to its development and evolution.
OpenStack architecture
OpenStack components
OpenStack is made up of a number of interconnected services that enable the automated creation, management and deployment of IT resources. The main components of OpenStack are :
- Nova: the compute service for creating and managing virtual machine instances.
- Swift: the object storage service offering a highly available, distributed storage solution.
- Neutron: the network service that lets you create and manage virtual networks and configure connections between different instances.
- Cinder: the block storage service for provisioning and managing storage volumes for instances.
- Keystone: the identity service that manages authentication, authorization and management of users and projects.
- Glance: the virtual image management service for storing, cataloging and sharing instance images.
- Heat: the orchestration service for deploying and managing complex infrastructures using user-defined templates.
- Horizon: the graphical user interface for visualizing and managing OpenStack resources.
Communication between components
The various OpenStack components communicate with each other via RESTful APIs (application programming interfaces). This enables services to exchange information and perform actions using standard HTTP requests. For example, when you create a virtual machine instance with Nova, this service communicates with Neutron to configure the network, with Glance to retrieve the instance image, and with Keystone to verify user authentication.
This modular architecture and the communication between components enable OpenStack to be highly flexible and extensible. Users can add or remove services as needed, and developers can create new services or customize existing ones using OpenStack's APIs.
Resource management with Nova
Nova component presentation
Nova is the central OpenStack component that manages compute resources, such as virtual machine instances. It enables instances to be created, deployed, scheduled and monitored on the various compute nodes.
With Nova, you can provision instances on demand, specifying their size, operating system and resource requirements. Nova then takes care of starting the instance on the appropriate host, and configuring the network to enable communication with other OpenStack instances and services.
Allocation of resources to bodies
In Nova, computing resources are managed from a resource pool. When you create an instance, Nova dynamically allocates the necessary resources from this pool, according to the specifications you have provided. This ensures efficient use of resources, avoiding waste and optimizing performance.
Quotas can also be defined to limit resource consumption per project or per user. These quotas can be adjusted according to the needs of each user or project, to optimize the use of available resources.
Planning and load balancing management
Nova also integrates advanced scheduling and load balancing functionalities. It can take into account various criteria, such as resource availability, system load and security policies, to determine on which compute node an instance should be deployed.
Nova also enables dynamic load-balancing mechanisms, automatically redistributing instances to different compute nodes according to system load. This guarantees a balanced use of resources, avoiding bottlenecks and ensuring high availability of services.
Storage with Swift
Introduction to the Swift storage service
Swift is OpenStack's object storage service. It enables the storage of large quantities of unstructured data, such as images, files, videos or documents. Swift offers a distributed, highly available and redundant storage solution, guaranteeing data resilience and security.
The Swift principle is based on the distribution of data over several storage nodes, called "rings". Each ring is responsible for a part of the storage and contains several physical disks. Data is replicated on different nodes to ensure availability in the event of hardware failure or breakdown.
How the distributed storage system works
When you store an object in Swift, it is fragmented into several parts and distributed across the various storage nodes in the ring. Each part of the object is replicated on several nodes, guaranteeing data availability and resilience.
When you need to access an object stored in Swift, it is reconstituted from the various parts present on the storage nodes. The transparency of this operation guarantees fast, easy access to data, even in the event of failure of one or more nodes.
Swift also offers advanced features such as data compression, metadata management and content indexing and retrieval. This makes it a powerful and versatile storage solution, suitable for a wide range of use cases.
Network with Neutron
Neutron basics
Neutron is OpenStack's network service. It lets you create and manage virtual networks, subnets, gateways and security policies. Neutron offers total flexibility in network configuration, allowing complex topologies to be defined and advanced security policies to be applied.
With Neutron, you can create isolated networks for each project or user, using private IP addresses. You can also define security rules to control data flows between individual instances and other networks.
Creating and managing virtual networks
In Neutron, you can create virtual networks by specifying their characteristics, such as IP address, subnet mask, gateway and security rules. Once a network has been created, you can attach subnets, gateways and virtual machine instances.
Neutron also offers the possibility of connecting virtual networks created in OpenStack with external physical networks, using virtualization mechanisms or hardware gateways. This makes it possible to extend OpenStack virtual networks beyond the platform's boundaries, connecting them to other existing infrastructures.
Routing and security configuration
Neutron lets you configure advanced routing rules to direct traffic between different instances and other networks. You can define static routes or use dynamic routing protocols, such as BGP (Border Gateway Protocol) or OSPF (Open Shortest Path First), to optimize network paths and guarantee high performance.
When it comes to security, Neutron offers advanced features for applying security rules at network, subnet or instance level. You can define access control lists (ACLs) to control incoming and outgoing data flows, and security groups to apply consistent security policies across multiple instances.
Orchestration with Heat
Introducing the Heat orchestration service
Heat is OpenStack's orchestration service. It lets you model and deploy complex infrastructures using user-defined templates. With Heat, you can create and manage resource stacks in an automated way, by specifying dependencies between different resources.
A Heat template is a text file in YAML (Yet Another Markup Language) format that describes the various resources and their properties, as well as the relationships between them. For example, you can define a template to create a stack containing a virtual machine instance, a storage volume and a network.
Infrastructure modeling and deployment
With Heat, you can model your infrastructure by describing the resources you need, their properties and relationships. You can define resources for virtual machine instances, storage volumes, networks, SSH keys, security groups and more.
Once your infrastructure has been modeled, you can deploy the stack using the Heat template. Heat then takes care of creating the requested resources, respecting the dependencies and constraints specified in the template. You can also update or delete an existing stack, by modifying the template or specifying the actions to be performed.
Heat also offers advanced features for error handling, event management and stack modification tracking. This enables efficient automation of infrastructure management and maintenance, guaranteeing consistent, reliable deployment.
Identity and security with Keystone
Roles and authorizations in OpenStack
Keystone is OpenStack's identity service. It enables authentication, authorization and management of users and projects. With Keystone, you can assign roles and permissions to users, to control their access to OpenStack resources.
In Keystone, users are associated with projects, which are isolated workspaces in which they can create and manage resources. Roles are used to define the permissions and privileges of users within a given project. For example, you can assign the role of administrator to a user, giving him or her full rights over the project.
User and project management
With Keystone, you can create, modify and delete users and projects, assigning roles and permissions. You can also set up security policies to control access to resources, specifying authentication rules and security levels.
Keystone also offers integration mechanisms with other authentication systems, such as LDAP (Lightweight Directory Access Protocol) or Active Directory, making it easy to integrate OpenStack into existing environments.
Dashboard with Horizon
OpenStack graphical user interface
Horizon is OpenStack's graphical user interface. It is a web-based dashboard that enables you to view and manage OpenStack resources in an intuitive and user-friendly way. Horizon offers a simplified user experience, enabling users of all skill levels to interact with OpenStack.
With Horizon, you can access OpenStack's various functionalities, such as creating and managing instances, networks, storage volumes, virtual images and more. You can also monitor instance performance, manage security policies, configure routing rules and more.
Key features and uses
Horizon offers a host of features that simplify OpenStack management and administration. You can create customized dashboards to organize and display the information most relevant to you. You can also set up alarms and notifications to keep you informed of important events.
Horizon also provides access to reports and statistics on resource utilization, enabling you to optimize resource allocation and control costs. You can also perform automatic actions using customized scripts and workflows, to automate repetitive tasks and boost efficiency.
Image management with Glance
Virtual image storage and management
Glance is OpenStack's virtual image management service. It enables you to store, catalog and share instance images, which serve as templates for the creation of new instances. With Glance, you can create, import, update and delete virtual images, guaranteeing their availability and integrity.
Virtual images are files or snapshots of virtual machine instances containing the operating system, applications and associated data. They can be used to create new instances quickly and efficiently.
Using images in OpenStack
In Glance, you can store images of various kinds, such as system images, application images or data images. You can also create images from storage volumes or snapshots of existing instances.
Once an image has been imported or created in Glance, you can use it to create an instance in Nova. When creating the instance, you simply specify the image you wish to use as a template. Nova then starts the instance using this image, and configures the operating system and associated applications.
Glance also offers advanced features such as image compression, image sharing between users or projects, metadata management and image search using filters. This makes it easier to manage and use images in OpenStack.
Queuing system with Zaqar
Message queuing and distribution
Zaqar is the OpenStack queuing service. It provides an asynchronous messaging solution for queuing and distributing messages between different OpenStack services. With Zaqar, you can communicate reliably and efficiently between different instances, even in distributed or large-scale environments.
When you queue a message with Zaqar, it is temporarily stored in a queue before being processed. Messages are then distributed to waiting services or instances, using publish/subscribe or routing mechanisms.
Integration with other OpenStack services
Zaqar is tightly integrated with other OpenStack services, making it easy to communicate and manage. coordination between different instances and services. For example, you can use Zaqar to send notification messages between different instances, to coordinate tasks between different instances or to synchronize status between different instances.
Zaqar also offers advanced features such as message persistence, priority management, message duplication and queue replication. This makes Zaqar a reliable and efficient messaging solution, suitable for a wide range of use cases.