Table of ContentsView in Frames

How data compression detects incompressible data

Incompressible data detection allows you to check if a file is compressible and for large size files, you can check if a compression group within a file is compressible. Allowing incompressible data to be detected saves the system resources used by inline compression trying to compress incompressible files or compression groups.

For files with size less than 500 MB, inline compression checks if a compression group can be compressed. If incompressible data is detected within a compression group, then a flag is set for the file containing the compression group to indicate that the file is incompressible. During subsequent compression attempts, inline compression first checks if the incompressible data flag is set for the file. If the flag is set, then inline compression is not attempted on the file.

For files with size equal to or greater than 500 MB, inline compression performs a quick check on the first 4 KB block of each compression group to determine if it can be compressed. If the 4 KB block cannot be compressed, the compression group is left uncompressed. However, if compression of the 4 KB block is successful, then compression is attempted on the whole compression group.

Postprocess compression runs on all files irrespective of whether the file is compressible or not. If postprocess compression compresses at least one compression group in an incompressible file, then the incompressible data flag for that file is cleared. During the next compression attempt, inline compression can run on this file to achieve space savings.

For more information about enabling or disabling incompressible data detection and modifying the minimum file size to attempt quick check on a file, see the sis config command man page.