Installation Assumptions OpenStack Compute has a large number of configuration options. To simplify this installation guide, we make a number of assumptions about the target installation. For a longer guide that discusses the basis for these assumptions, see the OpenStack Operations Guide. You have a collection of compute nodes, each installed with Fedora 18, RHEL 6.4, Scientific Linux 6.1 or CentOS 6 + CR distributions (continuous release ( CR ) repository). You have a collection of compute nodes, each installed with Ubuntu Server 12.04. You have a collection of compute nodes, each installed with openSUSE 12.3. There are also an OpenStack Install and Deploy Manual for Ubuntu and an OpenStack Install and Deploy Manual for RHEL, CentOS and Fedora and an OpenStack Install and Deploy Manual for openSUSE. Debian also has OpenStack support, but is not documented here. You have designated one of the nodes as the Cloud Controller, which will run all of the services (RabbitMQ or Qpid, MySQL, Identity, Image, nova-api, nova-network, nova-scheduler, nova-volume, nova-conductor) except for nova-compute and possibly networking services depending on the configuration. The disk partitions on your cloud controller are being managed by the Logical Volume Manager (LVM). Your Cloud Controller has an LVM volume group named "cinder-volumes" to provide persistent storage to guest VMs. Either create this during the installation or leave some free space to create it prior to installing nova services. Ensure that the server can resolve its own hostname, otherwise you may have problems if you are using RabbitMQ as the messaging backend. RabbitMQ is the default messaging back-end on Ubuntu RabbitMQ is the default messaging back-end on openSUSE Qpid is the default messaging back-end on Fedora. 192.168.206.130 is the primary IP for our host on eth0. 192.168.100.0/24 as the fixed range for our guest VMs, connected to the host via br100. FlatDHCP with a single network interface. KVM or Xen (XenServer or XCP) as the hypervisor. On Ubuntu, enable the Cloud Archive repository by adding the following to /etc/apt/sources.list.d/grizzly.list:deb http://ubuntu-cloud.archive.canonical.com/ubuntu precise-updates/grizzly main Before you run apt-get update and apt-get upgrade, install the keyring: sudo apt-get install ubuntu-cloud-keyring On openSUSE, use the Open Build Service repositories for Grizzly: # zypper ar -f obs://Cloud:OpenStack:Grizzly/openSUSE_12.3 Grizzly Ensure the operating system is up-to-date by running zypper update prior to the installation. Ensure the operating system is up-to-date by running yum update prior to the installation. Ensure the operating system is up-to-date by running apt-get update and apt-get upgrade prior to the installation. On RHEL (and derivative distributions) enable the RDO repository to retrieve OpenStack Havana packages: # yum install http://rdo.fedorapeople.org/openstack/openstack-havana/rdo-release-havana.rpm The Extra Packages for Enterprise Linux (EPEL) repository must also be enabled. The EPEL repository provides additional packages required by OpenStack: # yum install http://download.fedoraproject.org/pub/epel/6/i386/epel-release-6-8.noarch.rpm The RDO and EPEL release packages are not architecture dependent. This installation process walks through installing a cloud controller node and a compute node using a set of packages that are known to work with each other. The cloud controller node contains all the nova- services including the API server and the database server. The compute node needs to run only the nova-compute service. You only need one nova-network service running in a multi-node install, though if high availability for networks is required, there are additional options.