In order to make the most efficient use of your Backup Media (backup tapes), you want to eliminate the storing of applications and related data files that are already archived on their retail software distribution discs.  But, you need to be careful to specify all of the network locations that contain data created by your employees.  Miss one location and that data runs the risk of being lost forever.  File Investigator categorizes your files into specific categories that enable you to pinpoint which files are User Data and which files don't need to be backed up.

Object Level vs. Block Level De-Duplication

The latest methods for reducing the amount of data being backed up use Block Level De-Duplication.  In the past, de-duplication meant comparing hash codes of every file to ensure that no duplicate copies of files waste valuable backup storage space.  More recently, companies are starting to use Block Level De-Duplication to compare individual blocks of data within each file.  This method detects duplicate data within the same file and can also be applied to detect redundant data from one file to the next.  This is similar to compression algorithms that compare individual bytes (characters in words) to reduce the size of a file, but it's used on a block of bytes at a time.  A typical block size might be 4,096 bytes, to coincide with the typical cluster size on today's hard drives.  Unfortunately, most file formats don't store their data and data structures in 4,096 byte blocks, so the Block Level De-Duplication over laps the object borders in files and leaves room for improvement.  Changes made in just one of a file's objects can affect the block hash values of neighboring blocks if they overlap the borders of the changed object.

File Expander detects the locations and sizes of the individual data objects in files, which provides variable sized blocks that are Customized to the differing characteristics of each file.  Now, you can compare the hash codes of each object in a file, and stay within the boundaries of the objects.  When one object, in a file, is changed it doesn't effect the comparable hash codes of the neighboring objects, because the hash code blocks don't cross over object borders.  This provides more unchanged duplicate hash codes that can be eliminated from the backup process.  Using Object Level De-Duplication tailors the process to each file type, much like file compression techniques tailor their algorithms to specific file types.  For example, you wouldn't want to use the A-Law sound compression algorithm on a bitmap image, because it isn't as efficient on image files as the JPEG image compression algorithm.