Virtualization 101, From Data Center To Desktop
Virtualization Benefits Abound
you have established your company's needs, have determined the hardware required and loaded your virtualization product, the migration of existing servers can be accomplished in many ways. Various backup products resulting images, such as those created by Symantec's Ghost or Acronis True Image, can be converted to virtual machines that are ready to be transferred and placed into production. Many Data Centers perform live migrations converting hardware dedicated servers directly to Virtual Machine images that are then placed in production. Virtual Machines controlled by the same host machine are not restricted to the same subnet (or network group), as the use of virtual switches and VLAN tagging are supported by most implementations. This affords an even greater level of control in the flow of data and its access by other systems or users.
Virtual Machines can also be run from storage arrays that have high speed copper or fiber interconnects to the virtualization host servers. With the highly dynamic nature of today's storage arrays' space allocation and failover capabilities of their multiple controllers per unit, volume size adjustments can be made on the fly and availability is further enhanced when in real-time a failed controllers array is assumed by another functional controller, until repair can be made.
Templates can be made to create additional virtual machines for roll-out as needed. Entire software development environments can be rolled out and started at once from predefined templates.
At the heart of all of this would be your virtual machine infrastructure management and monitoring software. This software can be setup to control failover, load balancing, and much more. If a hardware server went down it could roll-out a VM from template or backup and put it on another hardware node (server instance) to assume the production responsibilities of the VM on the failed hardware. Should server resource usage be unusually high on a particular hardware node, the infrastructure software would move VMs off that node to another as needed to maintain production levels.In case of natural disaster and for business continuity, an entire server infrastructure could be kept in space at a remote network operations center (NOC) that would assume the functionality of the failed infrastructures role within minutes.
