Printable View
You all must be knowing that NAS and SAN are the best known approaches to store data in corporate networks. New technologies such as FCoE and iSCSI are also on the rise. All storage solutions bring advantages and disadvantages. Only those who know them, can create storage solutions for the future. But I don't know much about the NAS and SAN. I am here to get some information related to channel and IP networks, infrastructure, etc. So please provide some useful information regarding the topic.
Whether personnel data, information about customers in Customer Relationship Management (CRM) or CAD graphics developers offices, each company has to cope with an increasing amount of data manageable. Everywhere is saved and the number of stored, processed or to manage data and files is growing in the company and on servers, disk subsystems and tape drives hold after day critical data available day in order to request to access and edit. In companies it is important to keep the number of storage media as clear as possible. Where previously allowed the art Direct Access Storage Device (DASD) as a point-to-point connection between a server and a data memory, such as tape or disk subsystem, the storage of large amounts of data, walk the manufacturers and users are now other ways. DASD led to an escalating number of servers, the administrative burden became greater and greater, the increased costs for storage management. To counter this developed expert network storage technologies, storage, a separate block in the network design should be carried: SAN (Storage Area Network) and NAS (Network Attached Storage) are two opposite approaches.
Basically share the memory techniques in two ways: channel networks and IP networks. Sewer systems are designed to be tied to computer storage systems. They provide a fast connection is available that works like a separate channel. In sewer systems, memory media to send their signals simultaneously over a common connection, without them interfering with one another here. This type of network infrastructure is not as common as the IP networks. If a company wants to use such a sewer system, subject to additional costs because of this special hardware is required, must be designed for the extra workers to install and manage these devices. A well-known channel network is used in a SAN (Storage Area Network) is used. A SAN is working mostly with the Fibre Channel technology (FC). Using NAS systems (Network Attached Storage) data over IP networks can be saved.
For most of DASD storage devices using SCSI interfaces are connected. The SCSI method uses the client / server principle: Someone asks someone replies. Also describes the SCSI standard, the interfaces and the transmission rates that are possible: The last parallel SCSI standard is called "Ultra 320 SCSI", with the number 320 on the transmission speed of 320 MByte / s indicating that bus width is 16 bits. With SCSI, for example, the client sends a read request over the bus to a server, the server responds by sending the read file to the client. Depending on the transmission speed is faster or slower, the first standard had a transmission rate of five MByte / s, over a length of six meters were connected to eight devices. A SCSI adapter is now built into the device electronics and is usually for connecting tape drives and hard drives used. The current standard Serial Attached SCSI (SAS) is working in parallel SCSI interface, which replaced, as the limits of the speed achieved was almost. SAS uses a protocol based on the version of SCSI, the connection technology, however, parallel rather than serially, as with SCSI, the transmission rate is currently at three Gb / s. Data storage but can be transmitted not only through a SCSI cable or through a SAS cable. Via the Internet protocol (with NAS ) and the so-called Fibre Channel (SAN) is a possible transfer. When can the NAS storage devices directly to the local network ( LAN - Local Area Network) connection. SAN operates the server in a proprietary high-speed network, which is usually implemented via fiber optic connections.
A storage area network differs from a traditional LAN , using it only transports data, the mass memory are sent to one or those coming from and the path to a user are. A SAN is to say a secondary network , the local network of a company's borders, the connecting point is the server . In a SAN, the storage media such as tape drives and disk subsystems, not just devices that are connected to a server. A SAN is full of these mass storage devices, they are an integral part of such a network. While the described direct-attached storage, a point-to-point storage method between a server and a storage medium, takes place in a SAN connection for multiple servers to multiple storage over a private network.
One of the reasons for the development of SAN was the bandwidth. While NAS systems are connected to the local network and available bandwidth to provide data on SAN operates independently of a LAN, thus achieving a higher payload capacity. This is around 90 percent, which means that 90 percent of the data that are moved from a SAN, which stores data that is retrieved from the media or written. For NAS this payload is around 60 percent. In a SAN block-based data transfer, a block access, the computer calls to individual data blocks from a disk. In a file-based data query as to NAS is used, calls the computer requests files from all computer. A SAN is mostly through a Fibre Channel Infrastructure (FC) operated and consists of several components: the server control the request of clients, the Fibre Channel switch (FC switch) are connected to the servers and mass storage connected to the storage themselves serve as the location for all types of data A SAN is so similar to a backbone, but this is the only backbone designed to bring huge amounts of stored data to the users or transported from there to the mass memory. Most SANs use SCSI as a communication protocol based on Fibre Channel then as a transport protocol. Such storage backbone can work in different ways.
FC-AL is a bus topology. Servers and storage devices are connected over a virtual ring. The data exchange is possible in such a ring in one direction and only between two components. If two devices exchange data over such a ring, the other connected devices to wait until the bus is free again. In one of the rings can merge up to 128 devices, connected devices share the available bandwidth. By far the most common way to set up a SAN, the switched fabric technology. The switched fabric technology also serves as the strongest and most resilient among the safest option Fibre Channel approaches. Often my user switched fabric method when they talk about Fibre Channel. Central unit of the switched fabric topology is called a Fibre Channel switch . Unlike a hub that connects a switch connected servers and storage, not only because it can be Backplane (high speed), multiple connections between servers and storage at full bandwidth with want, they work just like conventional switches in local area networks. With FC switches can be cascaded and storage networks meshing. Normally, Fibre Channel, a technique for fiber optic cable to connect over copper cable is still possible, but sometimes fails to limit the length of the transmission medium. Since, in a Fibre Channel network via fiber cable to the FC switches should be connected to the server also, they need a special adapter so-called host bus adapter (HBA). These cards require the server typically a PCI-X slot with a 64-bit bus or a PCI Express version. For smaller SANs to provide FC switches with four to twelve ports, Enterprise devices can have up to 32 ports, and the so-called Directors, the counterparts to the core switches in local area networks can be equipped with up to 140 ports.