The acronym NAS stands for network-attached storage and, in its simplest form, that is precisely what NAS is: storage directly attached to an Ethernet network.
A major distinction between a NAS device and a traditional file server is the operating system. Traditional file servers use Novell, Linux, OS/2, Windows or Unix to offer file-based services out to the network. But where these file servers have other modules, code and features that do not directly relate to serving files, the NAS device is tuned for a specific function, and so it lacks other features, such as an application server or a directory server.
While the first NAS devices to appear on the market used dedicated silicon chips to handle their file serving tasks, today they use chips powerful enough to handle the rigours of a true dedicated network-attached file server.
The first NAS devices were aimed at the low end of the file-serving market. Companies such as Adaptec, with its Snap Servers, and Cobalt Networks created these NAS devices to appeal to small companies that needed to add some file storage quickly and easily without all the overhead of us administrator-types getting in the way and mucking things up.
These early NAS devices were single or dual hard disk devices with a network port in the back and an embedded operating system on a bit of flash (much like today's consumer-grade wireless routers) that allowed you to create a share and connect a bunch of PCs to it. They operated over TCP/IP high in the OSI model (Layer 7), they had their own file system (usually something Linux-based until the recent advent of the Windows Storage Server) and in general did not allow block-level access to the disks they contained (think being able to format a new hard drive).
Fast forward a few years. The idea caught on -- people liked the concept of just plugging a device into their Ethernet network and having storage available a few minutes later. Companies like NetApp capitalised on this popularity and created an "enterprise NAS" device, elevating the lowly consumer-grade product to a powerhouse of centralised file serving, having many disks installed using external disk shelves, each containing 10 to 14 disks. Today these NAS devices have evolved to include a credible, full-featured, performance-tuned operating environment and software stack geared towards serving up files to your networked computers. These systems include features that allow non-Windows computers to attach to the shares like NFS (Network File System) and added more and bigger networking ports, block-based storage and 2 Gbps and 4 Gbps Fibre Channel interconnects. All of this while maintaining the simplicity of operation that the idea of NAS has embodied from its inception.
About the author: Tory Skyers is a senior systems engineer for Prudential Fox & Roach Realtors, an independently owned and operated member of The Prudential Real Estate Affiliates Inc., based in Pennsylvania, U.S. He frequently speaks at conferences such as Storage Decisions and also contributes regularly to SearchStorage.com's blog called Storage Soup.
Go back to the beginning of the NAS handbook.
