Notes on Audit Data Compression

Top  Previous  Next

When you enable data compression for the Audit file, the record to be stored and the memo fields are compressed before being saved to disk and it is decompressed after being read from the disk. As the compression is done in memory, it is very fast, and as it reads and writes less data from and to the disk, it is possible to experience no performance delay during the auditing process. However, depending on your selected file driver storage, you may note a dramatic saving in the audit file size. You will have to experiment with your own configuration to see if this feature is a viable option or not.

 

The compression and decompression operations are performed by a open-source library called "zlib". You can download the latest version from www.zlib.net. For your convenience, FullRecord's setup installs the necessary libraries, which are zlib.dll, zlib.lib and zlibwapi.dll.

Note that when you use compression, even if you compile your project as local (Everything in the exe), zlib.dll will still be called by your exe, and it should be distributed with your program as an additional external dynamic link library.

 

For very small records the compression may produce undesired results, due to some header overhead information required by the zlib library. Our tests have shown that compressing records smaller than 40 bytes produced records that were actually bigger than the original size. However, in records over 200 bytes in length the compression ratio was over 50%. This is highly dependent on the type of data you are saving and results may vary from test to test.