First used by a company named Ardence (now functioning under Server leader Citrix), OS Virtualization is all about running a network set up on a virtual disk without preinstalling anything.
In an operating system (OS) virtualization, the local device does not require any hard disk as there is nothing loaded on to it. A virtual disk in the network is used to run all applications. This virtual disk is actually a disk image file that is stored on a remote server via a storage attached network (SAN) or a network attached storage (NAS). The local device being used by the client will boot through the OS on the virtual disk to which he is connected via the network. The kernel of the OS allows multiple isolated users to interface with it; instead of only one as is the case in a desktop computer.
Where is it used?
OS Virtualization is used where hardware resources are limited and where it is necessary to have a number of users each with his local device that is devoid of any hard disk. This greatly reduces hardware cost and allows the users to log onto a secure environment. The absence of the hard disk protects that device from any security problems.
It also enables applications to be segregated in different containers on the server, so that each client computer will be allowed to access only those applications that concern his particular needs. This greatly increases security and privacy and in corporate environments limits users to only those areas where they have a need to know.
Flexibility in OS Virtualization
The flexibility in an OS Virtualization is compromised when any of the clients require an OS that is different from the original OS. There has been an attempt to overcome this limitation in flexibility by the use of Solaris Containers which does allow different versions of the Linux OS to be used even where OS Virtualization is in place.
Storage in OS Virtualization
A private virtual disk is that which is used by one client only, and is very similar to a local hard disk. The user can save information on the virtual disk, and when the client restarts his settings are retained just as they would be on a local hard disk.
A shared or common virtual disk is used by a number of clients at the same time. Any changes made by the user to his data or files are saved in a special cache which gets automatically cleared as soon as that particular client shuts down. Every client has his own special identity which he would need to log onto the system.
How does OS Virtualization work?
The main item in the OS Virtualization infrastructure is the server. The server arranges to stream the information on the virtual disk to the client. It also determines which client is connected to a particular virtual disk. It hosts the storage itself or uses a storage attached network to do so. It ensures that each client has a unique identity throughout the connected infrastructure. The supporting components are a database that stores the configuration for the server and a streaming service for the virtual disk content. Optional are a TFTP (file saving protocol) service and a PXE boot service for booting.
The client when he boots his system through the virtual disk can then ask the server to connect him to the virtual disk which he is authorized to use and proceed to use the applications on it to complete his tasks.
The virtual disk will contain a physical image from the system that contains all the configurations and settings of the systems that will use the virtual disk. Once the virtual disk is created the connection to the client is made through the administrator who saves that particular clients settings in a database on the server or other storage systems linked to it.