At its most basic level, a storage area network (SAN) is simply a network dedicated to highspeed connections between servers and storage devices. Instead of installing disk drives into servers or connecting them by using an external SCSI bus, a SAN consists of one or more drive arrays equipped with network interface adapters, which you connect to your servers by using standard twisted pair or fiber optic network cables. A SAN-connected server, therefore, typically has at least two network adapters, one for the standard local area network (LAN) connection and one for the SAN, as shown in Figure 3-22.
The advantages of SANs are many. By connecting the storage devices to a network instead of to the servers themselves, you avoid the limitations imposed by the maximum number of devices you can connect directly to a computer. SANs also provide added flexibility in their communications capabilities. Because any device on a SAN can conceivably communicate with any other device on the same SAN, high-speed data transfers can occur in any of the following ways:
– Server to storage Servers can access storage devices over the SAN just as if they were connected directly to the computer.
– Server to server Servers can use the SAN to communicate directly with one another at high speeds to avoid flooding the LAN with traffic.
– Storage to storage Storage devices can communicate among themselves without server intervention, for example, to perform backups from one medium to another or to mirror drives on different arrays.
Although a SAN is not in itself a high-availability technology, you can make it one by connecting redundant servers to the same network, as shown in Figure 3-23, enabling them to access the same data storage devices. If one server should fail, another can assume its roles by accessing the same data. This is called server clustering.
Because they use standard networking technologies, SANs can also greatly extend the distances between servers and storage devices. You can design a SAN that spans different rooms, different floors, or even different buildings, just as you would a standard computer network.
Servers and storage devices cannot exchange SCSI commands over a SAN connection the way they do when the devices are directly connected using a SCSI cable. To communicate over a SAN, servers and storage devices map their SCSI communications onto another protocol, such as Fibre Channel.
Using Fibre Channel
Fibre Channel is a versatile SAN communications technology supporting various network media, transmission speeds, topologies, and upper-level protocols. Its primary disadvantage is that it requires specialized hardware that can be extremely expensive.
Note: FIBRE CHANNEL
The nonstandard spelling of the word fibre in Fibre Channel is deliberate, to distinguish the term from fiber optic. Fibre Channel can run on either twisted-pair copper cables or it can run on optical cables, whereas the spelling fiber always refers to an optical medium.
Installing a traditional Fibre Channel SAN entails building an entirely new network with its own special medium, switches, and network interface adapters. In addition to the hardware costs, which can easily be 10 times those of a traditional Ethernet network, there are also installation and maintenance expenses to consider. Fibre Channel is a rather esoteric technology, with relatively few experts in the field. To install and maintain a Fibre Channel SAN, an organization must either hire experienced staff or train existing personnel on the new technology. However, there is also a variant called Fibre Channel over Ethernet (FCoE) that uses standard Ethernet hardware and is therefore much less expensive.
Connecting virtual machines to a SAN
The specialized networking technologies used to build Fibre Channel SANs have, in the past, made it difficult to use them with virtualized servers. However, since the Windows Server 2012 implementation, Hyper-V has supported the creation of virtual Fibre Channel adapters.
A Hyper-V Fibre Channel adapter is essentially a pass-through device that enables a VM to access a physical Fibre Channel adapter installed in the computer, and through that, to access the external resources connected to the SAN. With this capability, applications running on VMs can access data files stored on SAN devices and administrators can use VMs to create server clusters with shared storage subsystems.
To support virtual Fibre Channel connectivity, the physical Fibre Channel host bus adapter(s) in the host computer must have drivers that explicitly support virtual Fibre Channel.
This support is relatively rare, but more manufacturers are expected to update their drivers to provide the necessary support. Your SAN must also be able to address its connected resources by using logical unit numbers (LUNs).
Assuming you have the appropriate hardware and software installed on the host computer, you implement the Fibre Channel capabilities in Hyper-V by first creating a virtual SAN by using the Virtual SAN Manager, accessible from Hyper-V Manager. When you create the virtual SAN, the World Wide Node Names (WWNNs) and World Wide Port Names (WWPNs) of your host bus adapter appear, as shown in Figure 3-24.
FIGURE 3-24 WWNNs and WWPNs in a virtual SAN
The next step is to add a Fibre Channel adapter to a VM from the Add Hardware page in the Settings dialog box. When you do this, the virtual SAN you created earlier is available on the Fibre Channel Adapter page, shown in Figure 3-25. Hyper-V virtualizes the SAN and makes the WWNNs and WWPNs available to the VM.
FIGURE 3-25 A Fibre Channel adapter in a VM
This article is a part of 70-410 Installing and Configuring Windows Server 2012 Prep course, more articles in this course are :
70-410 Installing and Configuring Windows Server 2012 Prep course includes following practice tests: