A journaling file system is a fault-resilient file system in which data integrity is ensured because updates to directories and bitmaps are constantly written to a serial log on disk before the original disk log is updated. In the event of a system failure, a full journaling filesystem ensures that the data on the disk has been restored to its pre-crash configuration. It also recovers unsaved data and stores it in the location where it would have gone if the computer had not crashed, making it an important feature for mission-critical applications.
Not all operating systems provide the same journaling technology. Windows NT offers a less robust version of the full system. If your Windows NT system crashes, you may not lose the entire disk volume, but you will likely lose all the data that hadn't yet been written to the disk prior to the crash. By the same token, the default Linux system, ext2fs, does not journal at all. That means, a system crash--although infrequent in a Linux environment--can corrupt an entire disk volume.
However, XFS, a journaling file system from Silicon Graphics, became a part of the open-source community in 1999 and, therefore, has had important implications for Linux developers, who previously lacked such insurance features. Capable of recovering from most unexpected interruptions in less than a second, XFS epitomizes the high-performance journaling filesystem of the future.
The earliest journaling file systems, created in the mid-1980s, included Veritas, Tolerant, and IBM's JFS. With increasing demands being placed on file systems to support terabytes of data, thousands upon thousands of files per directory and 64-bit capability, it is expected that interest will continue to grow in high-performance journaling file systems like XFS.