Understanding Virtual Networks
Virtual networks consist of virtual devices that are connected together in a virtual environment. Virtual machines (VMs), rendered by hypervisors, enable different VMs to share underlying hardware resources. Hypervisors (virtual machine monitors) are small-footprint software applications that reside between the hardware and the installed operating system, and sometimes behave as an extension of the BIOS. These applications divide the hardware resources (or host) into logical partitions (LPARs) and allocate dedicated or shared slices of these resources to each LPAR. Each VM (or guest) shares the physical resources of the host system. These include the CPU, memory, network interface card (NIC), and storage space. Figure 1 illustrates the logical partitioning of hardware resources into virtual devices.
Figure 1: Virtualized Server
Each VM is assigned a MAC address, is logically connected to virtual ports or switches, and initiates traffic from a virtual NIC. Virtual switches enables VMs on the different hosts to communicate with each other using the same protocols as physical switches. In addition, VMs can be configured with one or more virtual Ethernet adapters, each with its own MAC address and IP address. Virtual networks and virtual machines, therefore, have the same networking capabilities as physical networks built around physical devices.
The increase in the number of virtual switches in the data center presents Data Center (DC) operators with the challenge of having to manage these virtual network resources in addition to their physical networks. Network operators are unable to apply the tools that they currently use for managing the physical infrastructure to managing the virtual infrastructure.
