Brian Brannon, Director of Technical Operations
Recently, I have had repeated conversations with friends, family, and customers where I have been asked the same question over and over again. This question has been so prevalent in my day-to-day actions, that I feel that I am being targeted. Perhaps some secret society is testing my knowledge on this subject, and if I pass they will give me a large skeleton key and one million dollars. On second thought, maybe a rogue communist group is trying to determine which individuals need to be exterminated during their upcoming revolutionary coup. However, some could argue that I obviously watch entirely too much television.
What question is it that has gotten me into so much of a tizzy? Well, the answer to this question is “What is a SAN?” This question is very valid, and could have multiple meanings. For instance, San was a letter of the Greek alphabet between Pi and Qoppa. San could also be used as an abbreviation for the word sanitarium. Although the word or acronym “SAN” has many meanings, we will only discuss one of these. Our discussion will pertain to only SAN’s in the network infrastructure world.
First of all, SAN is an acronym that stands for Storage Area Network. However, this does not answer our overall question. The answer will take quite a bit more work and explanation. First we will need to address two other types of devices with their own acronyms that reside in the same network infrastructure realm as SAN’s.
The first of these devices that need an explanation is a DAS. DAS stands for Direct Attached Storage. This device is basically a rack mountable cabinet that contains a multitude of hard drives. On the back of a standard DAS is a SCSI connection that allows a single server, outfitted with a compatible SCSI card, to connect to the cabinet. This single server can then access the DAS as a native drive. Now, there is a little more intelligence and features built into some DAS’s, but overall this is the characteristics of a simple DAS.
The second device that warrants an explanation is a NAS. NAS stands for Network Attached Storage. Although a NAS is akin to the SAN, it is still a distant relative. In a sense, NAS’s are servers with large amounts of storage space. This storage space is accessible via file based protocols such as SMB and NFS. This means that folders will be created on the NAS, and shared out using one of these protocols. Once this is done, multiple servers can access these folders. However, this shared folder will be a network target and will not appear on the server as a local drive.
Now, what if you could take the technology of a DAS, where servers view the disk storage as a local drive, but you could do this over some type of network in kind of the same manner as a NAS. However, you could build a great deal more intelligence into this device to give you more options and redundancy. If you were to do all of this, you would have created a SAN. In the past SAN’s were only found in Fortune 500 companies. Recently however, SAN technology has dropped in price, and can now be found in a plethora of different sized entities. Hence the reason I am writing this article.
SAN’s are basically like a NAS in the fact that they are a server or servers that have many hard drives with a specialized operating system. However, SAN’s allow you to create different volumes that can be mounted on servers that will appear to be local drives. This is accomplished by accessing these volumes across a network using technologies like iSCSI or Fibre Channel. This allows companies to use clustering technologies to offer highly redundant systems.
Most SAN’s offer additional features and technologies that make them desirable to many individuals. For instance, snapshots allow one to take a picture of a SAN volume at a specified point-in-time. If something happens to the data on the volume, the snapshot can be mounted in an effort to revert back to that point-in-time. Also, data replication is a common feature of many SAN’s. Replication allows one to move data from one SAN, which reside in a datacenter, to another SAN that is located in a disaster recovery site. Thus, in the event of a disaster, a server at the disaster recovery site can mount the volume, and the data would be accessible instantly.
Virtualization is proliferating the use of SAN technology, and is a driving factor in the cost reduction trend that we see today. Virtualization is another discussion in and of its self, but all of the features, such as high availability, that make Virtualization tantalizing, require that a SAN be included in the Virtualization infrastructure. If someone tells you that you can obtain a highly available and redundant virtualized scenario without the use of SAN technology, you need to be asking questions about how this is being accomplished.
So, overall I believe that our primary goal of answering the question of “What is a SAN?” has been answered. Now, this article may very well have spawned off many other interesting questions about the use of a SAN; however, the answer to those questions could get a little dirty and technical. In the very near future, I believe that SAN’s will become more and more common in smaller networks that have been void of SAN’s up to this point. Until that time, I am sure that I will have to continue to answer the question of “What exactly is a SAN?”