BLENGOK

According to the Dictionary Industry Association storage networks (and who should know better?):

Storage Area Network (SAN) is a high-performance network, which is to enable the storage devices to communicate with computer systems and with each other primary purpose.

We believe that the things most interesting about this definition is not what he says:

It does not say that the only purpose for SAN is communication between computers and storage. Many organizations are working on a removable storage area networks completely to life, which sometimes carry traffic management application and others.

It does not say that the SAN uses Ethernet or Fibre Channel or any other specific link technology. There is a growing number of networking technologies have architectural and physical properties that make them suitable for use in storage areas.

He did not say what kind of interconnected storage devices. Whole disk and tape drives and subsystems RAID and robotic libraries, and file servers used fruitfully in SAN environments today. One of the exciting aspects of the SAN technology is that it is encouraging the development of new types of storage devices that provide new benefits for users. Some of these are undoubtedly in the market, but fail those that succeed will make lasting improvements in the way in which digital information is stored and processed.

Lets dig a little deeper in this definition.

What makes a different SAN?

Anyone in the field of information technology is already well known that the computers are connected to the storage devices. If this is the SAN everything he does, all that is new or different about it? The answer is simple, we will keep coming back to again and again throughout this book:

Global Storage Connectivity

Is connected computers to actually store the day, but all of the computers to install and connect all of its storage? This is the fundamental point about the storage-area networks come into contact with a lot of computers to a lot of storage devices, enabling computers to negotiate device ownership among them and, ideally, to exchange data. If there is one hallmark of SANs, its global connectivity to storage devices and computers.

To estimate the value of the storage universal plug, look at the client computer / traditional server system shown in Figure 1.1.

Figure 1.1

Figure islands of information / Server 1.1 client

From this image, business as usual computing client / server, it is immediately apparent that through the deployment of multiple servers, an organization that creates an interconnected islands of information automatically. Each island can be accessed via a single computer, but not others. If computer B needs to use the data produced by computer A, the data has to be copied to computer B.

There are many techniques to transfer data between computers: backup, file transfer. And communication between processes, for example, is not limited. But the real issue is that the organization of information services have to own and manage the additional resources required both to copy the data from computer to computer B and stored in both locations. There is no reason for this to work duplication of effort, and others that the computer needs data that has been produced by another computer.

There are more serious implication of information which depend on regular copying of data from one computer to another treatment strategy. Often forced computers that access to copy the data to work with the data outdated simply because it is physically impossible to make a copy in a timely manner. Moreover, the additional operational complexity introduced by the need to copy data between servers creates an additional opportunity to expensive mistakes.

By contrast with the system architecture and distributed based SAN shown in Figure 1.2.

Figure 1-2

Figure 1.2 SAN eliminates islands of information

With SAN, and the concept of a host computer and one that has the data stored or not meaningful. All servers in Figure 1.2 of a physical link to all storage devices. If F server needs the data that was produced by E server, there is no need to copy, because the server F can be accessed on any hardware E Data Storage Server. All that is required is a change of ownership of logical storage device from server to server EF or better yet, to pre-E server agreement to stay out of the way while actively use the data server F.

Mass storage unit connection has some very powerful effects circuits information services:

Theres no need to set a timetable middle of the data transfer between pairs of servers.

Theres no need to purchase and maintain additional storage to organize one server data temporarily in another server.

Theres no need to worry about whether they are synchronized copies of the data that is used by the two computers running different applications (which is exactly the same content), because the computers are working from the same copy of the data.

In fact, at this level simplistic, it is difficult to see how any organization responsible for electronic information processing can not be that he wants SAN. Lets dig deeper still and see how true this is.

What makes a good SAN?

SAN architecture completely interconnected appear on the right side of Figure 1.2 intuitively attractive, but if I want her to be the backbone of I / O to process information services, it needs a couple of qualities:

There must be a SAN is available too. SAN and one connecting all the computers for all storage puts a lot of eggs access to company information in one basket. SAN to be the best non very destruction or institution can be literally out of business. SAN good implementation will have built-in protection from just about any type of failure imaginable. As we will see in later chapters, this means that not only must the links and keys infrastructure consists SAN be able to survive the failure element, but storage devices, and interfaces to the tongue, and computers themselves, and we must all have built-in strategies to survive and recover from failures also .

I / O performance for SAN must grow or scale as the number of connected devices is growing. If SAN connects a lot of computers and plenty of storage, it may be better able to deliver the performance they all need to do their jobs, including one at a time. The best SAN delivers both high and low data I / O request latency transfer rates. Moreover, it must be the performance of SAN and able to grow as the organization of information and processing needs grow storage. As is the case with the networks of other companies, it is not just a process to replace SAN often.

On the positive side, a SAN to do the scope of the additional application and provides enhanced performance by separating Size-high I / O traffic from the message traffic client / server, and give each track is optimized for its properties and eliminate cross-talk between them.

Investment required to implement high-SAN, both in terms of direct cost of capital or in terms of the time needed to learn the technology and design, deploy, and melody, and SAN management and energy. Any institution is well managed to do cost-benefit analysis before deciding to implement storage networks. The results of this analysis suggests it is almost certain that the largest yield comes from the use of SAN to connect the most important enterprise data to computers that are running the most important applications.

But the most important data is the data that can least afford institution to be without. Together, the natural desire to maximize the return on investment and the importance of operational data lead to Article 1 of the storage networking:

When designing SAN access to critical corporate data, make sure SAN available too (ie can survive the failure of all of the ingredients in it, and components associated with it) and make sure it can grow beyond the needs peak expected performance without interruption.

What Makes a Great SAN?

Therefore, a key feature of the storage areas are the global connectivity data storage. Comprehensive connectivity enables a range of important advantages. Depending on the particular SAN hardware and software components selected, additional benefits may be realized from the advanced functions are being built in the SAN devices today. Again, we describe some of the features and benefits identified in later, but for now affirm Article 2 of storage networks:

When assessing the implementation SAN options, once can be achieved capacity, availability, and the requirements of the basic performance, and the search for advanced functions available in the architecture selected and look at how they can be used to further reduce the cost or enhance the information provided to users services.