What Is a Grid File System?

Eugene P.

A grid file system is a type of computer architecture designed to combine several storage devices in a way that allows them to be accessed as a single logical drive by multiple users. Both the storage devices and the users can be located anywhere, from within the same room to across the world, and the grid file system will optimally abstract the interfaces so all users are interacting with the system in a consistent way. A properly implemented grid file system is one component of a fully distributed computing system in which many, or all, resources are shared among users. In addition to sharing a single virtual drive, a grid file system also can implement redundancies to protect data from catastrophic loss and load balancing to ensure that the grid file system is available at all times. Some difficulties in creating a grid file system include preventing latency from corrupting data, providing a fast and secure client-server setup for file access, and ensuring that a common namespace is maintained and propagated through all independent networks using the file system.

Grid file systems are used to prevent catastrophic data losses and to ensure the system is available at all times.
Grid file systems are used to prevent catastrophic data losses and to ensure the system is available at all times.

The concept of a grid file system is the realization of distributed resources across a network. The ultimate goal is to allow any user to access the grid system in a natural way and be able to interact with files as if they were on a local disk. All users should be able to use the exact same file and directory structure, regardless of where the data is located. This means that, with redundancy, two users might be looking at the same file but the file loaded and displayed in each case might have been loaded from a different physical disk depending on the availability of the drives at the time of the request. The mechanics of the file system make this type of interaction transparent to the users.

One large complication with using a grid file system is latency. A situation can arise in which two users are looking at the same file but the file is spread across multiple drives in both cases. If one user changes the file and saves it within the same time frame as another user, unpredictable results might occur. Depending on several factors, two different versions of the file might exist in two locations within the grid, or only one file might be saved, depending on the implementation of the file system and the hardware.

Complex problems, especially with wide-area networks (WAN) and Internet-based grids, have prevented any type of standardization in the development of grid file systems. Grids that are built tend to be custom made, along with custom software solutions. For this reason, the use of a formal grid type file system has been restricted mostly to non-critical uses and research.

You might also Like

Discuss this Article

Post your comments
Forgot password?