What is file virtualization?

File virtualization is a type of storage virtualization that occurs on the file-level. Storage virtualization involves abstracting logical storage (virtual storage) away from the physical storage (storage hardware). The logical storage is generated from the physical storage, which is combined into storage pools. File virtualization is a vital part of network file management (NFM) and file area networks (FANs).

There are two methods of file virtualization: built-in and standalone. The first method has virtualization built into the NAS (Network Attached Storage) or storage system, creating a ‘global’ file system. This file system manages the metadata, such as file stubs, for file locations. A disadvantage of this method is that it locks you in to the hardware you’re currently using because the metadata is exclusive to the users of the file system. To expand these systems, you will typically have to buy hardware from the same vendor, so you give up your ability to mix various brands of hardware. Global file systems also don’t have file-level granularity in many cases.

The other method for file virtualization employs a standalone virtualization engine or appliance. These engines can be located within the data path (in-band) or outside of the data path (out-of-band). In-band engines often have better performance, but out-of-band engines are easier to set up. Standalone solutions move files based on attribute related policies and have file-level granularity. They also allow you to use any combination of hardware.

Benefits
File virtualization reaps several benefits that are found in every virtualization strategy, including unified/simplified management, more efficient use of storage capacity, shielding users and administrators from the internal complexity of the storage environment, and reduced cost of management.

There are also a number of benefits that are specific to file virtualization. It eliminates the need for a manager to know a file’s physical location on the system. Instead, there’s a single point where files can be accessed. File virtualization allows storage managers to move data around at will. You won’t need to worry about changing users’ logins if, for example, you buy a new file server or NAS and have to migrate data from an older system. Some file virtualization software can even transfer data in the background so that new storage can be added without causing problems for the people currently using your application.

Additionally, file virtualization enables the creation of a tiered storage or information lifecycle management (ILM) strategy. The ability to move data between one to three-tiered NAS/network file servers allows files to be moved into more appropriate systems based on how frequently they are accessed. This can minimize overhead and operational costs. Accessing older data in a tiered storage system is also much easier to manage with file virtualization because there’s no need to create and keep track of a stub file to access the data.