We'll run through an example of installing a CentOS image. This will focus mainly on CentOS 6.4. Because the CentOS installation process may change across versions, if you are using a different version of CentOS the installer steps may differ. Navigate to the CentOS mirrors page. Click one of the HTTP links in the right-hand column next to one of the mirrors. Click the folder link of the CentOS version you want to use. For example, 6.4/. Click the isos/ folder link. Click the x86_64/ folder link for 64-bit images. Click the ISO image you want to download. The netinstall ISO. For example, CentOS-6.4-x86_64-netinstall.iso is a good choice since it's a smaller image that will download missing packages from the Internet during the install process. Start the installation process using either virt-manager or virt-install as described in the previous section. If using virt-install, don't forget to connect your VNC client to the virtual machine. We will assume the name of your virtual machine image is centos-6.4, which we need to know when using virsh commands to manipulate the state of the image. If you're using virt-manager, the commands should look something like this:# qemu-img create -f qcow2 /tmp/centos-6.4.qcow2 10G # virt-install --virt-type kvm --name centos-6.4 --ram 1024 \ --cdrom=/data/isos/CentOS-6.4-x86_64-netinstall.iso \ --disk /tmp/centos-6.4.qcow2,format=qcow2 \ --network network=default \ --graphics vnc,listen=0.0.0.0 --noautoconsole \ --os-type=linux --os-variant=rhel6 At the initial Installer boot menu, choose the "Install or upgrade an existing system" option. Step through the install prompts, the defaults should be fine. The default TCP/IP settings are fine. In particular, ensure that Enable IPv4 support is enabled with DHCP, which is the default. Choose URL as the installation method. Depending on the version of CentOS, the net installer requires that the user specify either a URL, or the web site and a CentOS directory that corresponds to one of the CentOS mirrors. If the installer asks for a single URL, an example of a valid URL would be: http://mirror.umd/centos/6/os/x86_64. Consider using other mirrors as an alternative to mirror.umd.edu. If the installer asks for web site name and CentOS directory separately, an example would be: Web site name: mirror.umd.edu CentOS directory: centos/6/os/x86_64 See CentOS mirror page to get a full list of mirrors, click on the "HTTP" link of a mirror to retrieve the web site name of a mirror. If asked about what type of devices your installation involves, choose "Basic Storage Devices". The installer may ask you to choose a hostname. The default (localhost.localdomain) is fine. We will install the cloud-init package later, which will set the hostname on boot when a new instance is provisioned using this image. There are different options for partitioning the disks. The default installation will use LVM partitions, and will create three partitions (/boot, /, swap), and this will work fine. Alternatively, you may wish to create a single ext4 partition, mounted to "/", should also work fine. If unsure, we recommend you use the installer's default partition scheme, since there is no clear advantage to one scheme or another. Step through the install, using the default options. The simplest thing to do is to choose the "Basic Server" install (may be called "Server" install on older versions of CentOS), which will install an SSH server. Once the install completes, you will see the screen "Congratulations, your CentOS installation is complete". To eject a disk using virsh, libvirt requires that you attach an empty disk at the same target that the CDROM was previously attached, which should be hdc. You can confirm the appropriate target using the dom dumpxml vm-image command. # virsh dumpxml centos-6.4 <domain type='kvm'> <name>centos-6.4</name> ... <disk type='block' device='cdrom'> <driver name='qemu' type='raw'/> <target dev='hdc' bus='ide'/> <readonly/> <address type='drive' controller='0' bus='1' target='0' unit='0'/> </disk> ... </domain> Run the following commands from the host to eject the disk and reboot using virsh, as root. If you are using virt-manager, the commands below will work, but you can also use the GUI to detach and reboot it by manually stopping and starting.# virsh attach-disk --type cdrom --mode readonly centos-6.4 "" hdc # virsh destroy centos-6.4 # virsh start centos-6.4 In theory, the virsh reboot centos-6.4 command can be used instead of using destroy and start commands. However, in our testing we were unable to reboot successfully using the virsh reboot command. When you boot for the first time after install, it may ask you about authentication tools, you can just choose "Exit". Then, log in as root using the root password you specified. An instance must perform several steps on start up by interacting with the metadata service. For example, retrieve ssh public key and execute user data script. There are several ways to implement this functionality, including: Install a cloud-init RPM, which is a port of the Ubuntu cloud-init package. This is the recommended approach. Modify /etc/rc.local to fetch desired information from the metadata service, as described below. The cloud-init package will automatically fetch the public key from the metadata server and place the key in an account. You can install cloud-init inside the CentOS guest by adding the EPEL repo:# rpm -Uvh http://download.fedoraproject.org/pub/epel/6/x86_64/epel-release-6-8.noarch.rpm # yum install cloud-init The account varies by distribution. On Ubuntu-based virtual machines, the account is called "ubuntu". On Fedora-based virtual machines, the account is called "ec2-user". You can change the name of the account used by cloud-init by editing the /etc/cloud/cloud.cfg file and adding a line with a different user. For example, to configure cloud-init to put the key in an account named admin, edit the configuration file so it has the line: user: admin If you are not able to install the cloud-init package in your image, to fetch the ssh public key and add it to the root account, edit the /etc/rc.local file and add the following lines before the line “touch /var/lock/subsys/local” if [ ! -d /root/.ssh ]; then mkdir -p /root/.ssh chmod 700 /root/.ssh fi # Fetch public key using HTTP ATTEMPTS=30 FAILED=0 while [ ! -f /root/.ssh/authorized_keys ]; do curl -f http://169.254.169.254/latest/meta-data/public-keys/0/openssh-key \ > /tmp/metadata-key 2>/dev/null if [ \$? -eq 0 ]; then cat /tmp/metadata-key >> /root/.ssh/authorized_keys chmod 0600 /root/.ssh/authorized_keys restorecon /root/.ssh/authorized_keys rm -f /tmp/metadata-key echo "Successfully retrieved public key from instance metadata" echo "*****************" echo "AUTHORIZED KEYS" echo "*****************" cat /root/.ssh/authorized_keys echo "*****************" done Some VNC clients replace : (colon) with ; (semicolon) and _ (underscore) with - (hyphen). Make sure it's http: not http; and authorized_keys not authorized-keys. The above script only retrieves the ssh public key from the metadata server. It does not retrieve user data, which is optional data that can be passed by the user when requesting a new instance. User data is often used for running a custom script when an instance comes up. As the OpenStack metadata service is compatible with version 2009-04-04 of the Amazon EC2 metadata service, consult the Amazon EC2 documentation on Using Instance Metadata for details on how to retrieve user data. In order for the instance to access the metadata service, disable the default zeroconf route: # echo "NOZEROCONF=yes" >> /etc/sysconfig/network In order for nova console-log to work properly on CentOS 6.x, guests you may need to add the following lines to /boot/grub/menu.lstserial --unit=0 --speed=115200 terminal --timeout=10 console serial # Edit the kernel line to add the console entries kernel ... console=tty0 console=ttyS0,115200n8 From inside the instance, as root:# /sbin/shutdown -h now The operating system records the MAC address of the virtual ethernet card in locations such as /etc/sysconfig/network-scripts/ifcfg-eth0 and /etc/udev/rules.d/70-persistent-net.rules during the instance process. However, each time the image boots up, the virtual ethernet card will have a different MAC address, so this information must be deleted from the configuration file. There is a utility called virt-sysprep, that performs various cleanup tasks such as removing the MAC address references. It will clean up a virtual machine image in place:# virt-sysprep -d centos-6.4 Now that the image is ready to be uploaded to the Image Service, you no longer need to have this virtual machine image managed by libvirt. Use the virsh undefine vm-image command to inform libvirt.# virsh undefine centos-6.4 The underlying image file you created with qemu-img create. For example, /tmp/centos-6.4.qcow2 is now ready for uploading to the OpenStack Image Service.