Network file management addresses the problems inherent in providing access to computer files that might exist across an array of network-connected devices. Typically, this involves some form of storage virtualization that enables a user to access files as if they were stored on his or her local machine. In a distributed file system, files may reside on one or more servers or network-attached storage (NAS) devices. A virtual user interface is implemented by means of a uniform file-naming system and the mapping of logical groups, such as directories, to locations in physical storage.
In most network file management scenarios, computers are either clients or servers. Clients are workstations that run applications and rely on servers for resources such as storage. Typically, clients do not have direct access to data, but interact with the server by means of a network file system (NFS) protocol. Access to files and permission to modify them may be subject to the user's level of security authorization. A virtual file system (VFS) overlays the protocol, allowing users access to the shared network files as if they were stored locally.
Specifically designed to store computer files, the NAS device has come to replace general-purpose computers often assigned to that task. Its dedicated design simplifies network file management and increases the efficiency of data delivery to workstations. The NFS server still mediates requests for service, but the NAS device is responsible for forwarding data and updating modified files. Since it is not a part of the server, multiple NAS devices can reside anywhere in the network and may be added or removed without shutting down the server.
These devices may also incorporate a further storage virtualization called a Redundant Array of Independent Disks (RAID). Files can be stored and copied across multiple disks inside the NAS while being accessed as if they were on a single storage medium. This is another scheme designed to make the complexities of network file management easier to administer. The user remains shielded behind a virtual interface and continues to work as if the data required was on hand, stored in his or her local machine.
Available physical storage in a network is divided into basic equal sized amounts called physical extents. These are mapped to units of virtual storage, called logical extents, of the same size from which groupings such as files, directories and volumes can be assembled. The location of the actual physical extents holding data is invisible and unimportant to the user. Virtual file systems mediate all requests to access and save data. The ability to work with storage on a virtual level simplifies network file management chores such as backup, security access and the ability to easily add or remove data storage resources.
