What Is Storage Area Network or SAN?
In computing, a storage area network, or SAN, is a specialized network to pool storage resources.
Storage Area Network – Definition
A SAN (or Storage Area Network) is different from other storage such as NAS (network attached storage) with a low level access to disks. For simplicity, the traffic on a SAN is very similar to the principles used to use internal drives (ATA, SCSI). This pooling of storage resources.
In the case of NAS, the storage resource is directly connected to the corporate IP network. The NAS server integrates support for multiple network file systems, such as Common Internet File System (CIFS) protocol shares of Microsoft and Samba, Network File System (NFS) is a protocol for sharing files on Unix, or even AFP (AppleShare File Protocol) which is equivalent in the Apple world. Once connected to the network, it can act as multiple servers to shared files.
In the case of SAN storage arrays do not appear as shared volumes on the network. They are directly accessible by the block device file system servers. That is, every server sees the disk space on a SAN array which is accessed as its own hard drive. The administrator must define precisely the Logical Unit Number (LUN LUN), the masking and zoning for a Unix server does not access the same resources that a Windows server using a different file system.
Storage Area Network – Benefits
An early interest in the SAN is not having to worry about how much to change discs once vested in a particular server or grieve the unused space on a server oversized disk space.
Disk space is more limited by the characteristics of servers and is scalable at will by adding disks or storage arrays on the SAN. The physical storage space for shared servers is used to optimize the management of records, and make easier data backups.
Storage resources and pooled (SAN or NAS) give the possibility to implement replication functions (copy data synchronously or asynchronously between two bays) and snapshot (copying a volume for use on another server or save for example). These functions allow secure data (physical location in remote locations) and to optimize application availability. These functions are carried out transparently for servers, and replication and copying data do not affect server resources, since they are performed at SAN controllers, the impact on response time is usually negligible.
Remote replication: SAN solutions have some ability to transfer data remotely, typically on a remote site as part of a backup plan.
Storage Area Network – SAN Features
Quality of Service (QoS)
The switch guarantees a fixed rate of 8 Gbit / s (above 1 Gbps, 2 and 4) optical fiber link, and ensures that the request sent by a server has been well received and taken into account by the storage.
The SAN can provide storage redundancy, that is to say, the accessibility to the storage system in the event of failure of one of its elements, at least doubling each element of the system (high availability).
The SAN can operate in a completely heterogeneous servers Unix, Windows, Netware … can all join the SAN.
The performance of disk access by a host of stress varies from the other hosts on shared resources, and thus may vary over time.
Types of SAN (Storage Area Network)
SANs are built in order to provide fast and reliable disk space. The most widespread technology to achieve this is to use the Fibre Channel protocol, which achieves high throughput (8 Gbps or 10 Gbps). This technology leads the implementation of switches or directors of (wholesale switches of more than 64 ports, guaranteeing maximum availability and scalability by adding cards containing data ports). The storage network thus formed is called a fabric.
Other technologies have emerged:
- ISCSI: it is the same SCSI commands but TCP / IP;
- FCoE (Fibre Channel over Ethernet) and Fibre Channel protocol frames are transmitted over an Ethernet network.
The most used protocol and standardized remains the Fibre Channel.
The main difficulty in the design and operation of a SAN is to ensure to every development of the compatibility of different components that are implemented: Fibre Channel cards with their version of firmware, driver (driver) software multipath prospective switches firmware, version controller storage arrays. It is essential to respect the recommendations of the manufacturer or integrator (compatibility matrix) to ensure proper operation and performance of applications hosted on the SAN.
Storage Area Network – SAN action
There is still some time we could find that SANs in large computer companies in need of a large volume of highly secure storage. Prices becoming more reasonable, smaller data center equip themselves with SAN architectures simpler.
A larger site will create a more complex architecture composed of several interconnected SANs, each SAN may be equipped with a large number of arrays may be heterogeneous.
The identification of all the physical parts of a SAN is done through the World Wide Name (WWN), a unique identification of the equipment.
All switches (switches) connected together is an object named virtual fabric. In a single fabric, zoning information (permissions) are known to all switches.
The masking, meanwhile, is an action performed on the storage array and which aims to correlate the access permissions to logical drives created.
The SAN is also a big risk for DBMS and needs IO bases must be finely measured. Some publishers like Microsoft Exchange recommend dedicating SAN them exclusively.
Study: From Wikipedia, the free encyclopedia. The text is available under the Creative Commons.
- Cloud Computing: The Concept and Examples of its Virtual Services | Part 1 - July 23, 2012
- Why Rapidly Growing Companies Need Cloud Computing | Part 1 - July 22, 2012
- Web Designing Process | Strategic Planning | Part 1 - August 7, 2011