Important Pre-Installation Step

Make sure that you open the ports 8888 and 8887 in the firewall. Otherwise you won’t be able to connect to HyperVM.

You have to disable selinux by editing /etc/sysconfig/selinux and changing the line to selinux=disabled and then running
# $ setenforce 0

HyperVM installation consists of downloading either the hypervm-install-master.sh or hypervm-install-slave.sh from http://download.lxcenter.org and execute it as root. It will download all the required files and do the complete installation on it’s own.

Install Commands

# $ setenforce 0
# $ wget http://download.lxcenter.org/download/hypervm/production/hypervm-install-master.sh
# $ sh ./hypervm-install-master.sh --virtualization-type=xen/openvz/NONE

You have to explicitly specify which virtualization type you want using the option –virtualization-type as either xen or openvz or NONE.

If you want to install hypervm on an existing system, just use virtualization-type=NONE, which will install just hyperVM components, and skip the virtualization components entirely.

When you run the command, you will presented with a menu from up2date, on which you can press .

HyperVM will currently install the openVZ/Xen kernel, but will not switch the default kernel to it. You will have to manually edit the /etc/grub.conf, and change the ‘default=1′ to ‘default=0′, and reboot the machine.

HyperVM supports the concept of multiple locations, and thus generally the question of harddisk partitioning is irrelevant. It is recommended to have a large / for openvz, since both /home and /vz will have files. You can just add another harddisk, mount it as a particular directory, and then add it to the list of locations. In other words, you can trivially distribute vpses across different harddisks/lvms. Also you can change the location of a vps from one harddisk to other from the CP itself. Go to vps home -> advanced -> change location.

You should never run any other control panel other than hyperVM on the host server. That will unncessarily introduce security risks, and in the case of openvz, since all the vps process are visible and accessible on the host server, starting and stopping of services on the host server will affect all the vpses too. If you need to run anything, create a special vps for it.

Once the machine boots back, just make sure you have the openVZ kernel by doing a ‘uname -a’, and after that you can connect to ‘http://machine-name:8888′, and you will be presented with a login screen – the password would be ‘admin’. Once you login, HyperVM will explicitly force you to change the password to something other than ‘admin’.

For those installing Xen, please read through the forum notes, and make sure that you meet all the requirements.

If you are installing hyperVM on a legacy operating system (Fedora 2, 3, 4), in an existing virtualization system, there are some things that you have to take care:

* Move /usr/bin/yum to /usr/bin/yum.old . HyperVM cannot use the old yum owing the fact that its configuration files are different from the latest version. If you move yum to yum.old, hyperVM will automatically use up2date.

* You can only install hyperVM, and installing of virtualization is not supported on legacy platforms. The legacy support is primarily meant to help existing installations to migrate to hyperVM. New installations are supported only on Centos-4.4 for openvz and fedora -6 for xen. So only –virtualization-type=NONE will work on the legacy platforms.

If you want to have a cluster, you have to run ‘hypervm-install-slave.sh’ in all the servers that you intend to make as slaves. Once ‘hypervm-install-slave.sh’ is completed, you can add the server from the ‘add server’ page in the master’s server page.

You will need to open port 8889 which is what’s used for communication between the master and slave. For file transfer, the slave will need to contact the master’s 8889 port, so you will need 8889 open on both master and the slave. All communication is fully encrypted and secure.

From then on, you can manage every aspect of Virtualization through our intuitive graphical interface.