The OpenStack project is an open source cloud computing platform for all types of clouds, which aims to be simple to implement, massively scalable, and feature rich. Developers and cloud computing technologists from around the world create the OpenStack project.

This guide is written so that you can choose your own OpenStack adventure. Because OpenStack is highly configurable to meet different use cases, with many storage choices and underlying software-defined networking scenarios, this guide offers these main examples with the intention of expanding on each: Example architecture for simplicity: the most basic architecture you can have: one node to run the "control" services, such as database, message queue and API services for the Identity Service, Image Service and Compute Service; and a second node to be our Compute node that runs the hypervisor where our virtual machines will live.

Technical details: Compute with KVM, local ephemeral storage, nova-networks in multi-host flatDHCP, MySQL, nova-api, default scheduler, Qpid for messaging, Identity with SQL backend, Image with local storage, Dashboard (optional extra). Use as many default options as possible. Technical details: Compute with KVM, local ephemeral storage, nova-networks in multi-host flatDHCP, MySQL, nova-api, default scheduler, RabbitMQ for messaging, Identity with SQL backend, Image with local storage, Dashboard (optional extra). Use as many default options as possible. Example architecture from the OpenStack Operations Guide: Same as above but with Block Storage backend LVM/iSCSI, nova-network in multi-host with FlatDHCP, Live Migration backend shared storage with NFS, and Object Storage. One controller node and multiple compute nodes. Example architecture with Identity and Object Storage only (no Dashboard): Five node installation with Identity on the proxy node and three replications of object servers. Dashboard does not support this configuration so examples are with CLI. Example architecture with OpenStack Networking (Neutron)