The cloud in general is definitely one of the most discussed and interesting topics that come up in almost every discussion or meeting with IT folks.
During the last year OpenStack has been practically unknown for me so this summer I decided to play around with it in my lab.
After reading the official documentation I started installing OpenStack from scratch, using the minimal configuration. I chose CentOS because you can install the entire platform using RDO packages (OpenStack RPM distribution) but even if the packages are ready to be installed, with this process, you still need to customize the entire installation by editing hundreds of configuration files.
In my opinion, while with this process you can understand how OpenStack works and learn some useful CLI commands, on the other hand it will take you a lot of time to deploy your installation.
Here comes the RDO Community, which is a community of users who install, configure and manage OpenStack on Red Hat based distros.
Included in RDO packages there is Packstack, an utility that uses Puppet modules to deploy OpenStack on multiple servers over SSH. The installation process is controlled by an answer file, and with this file is possible to create almost every kind of scenario: a single node installation or a more complex scenario.
In this following post I will share the steps I followed to deploy my installation based on three nodes: controller, compute and network.