Monday, November 26, 2007

Benchmarks*: JkDefrag 3.28

You can find a review of the GPL freeware JkDefrag program (JKD), as well as the benchmark results obtained. You can download this version in an easy setup format. The version used for testing is 3.28, but newer ones are available.
The graph shows JkDefrag 3.28 (JKD) in dark blue, and the results from Windows XP in light blue. The first test is at the bottom, and shorter lines mean faster times. The magenta lines are the results obtained after running JKD from a BartPE environment, which allows the page file and other system files to be moved.
  • "Basic XP" refers to the standard install, without Office 2007, so there are only 710 files to be tested.
  • "Defrag" refers to the read times of the same 710 files, after two defrag passes using the JKD default defrag, with no "Optimize Files" or "Optimize Boot" options enabled. A 15% performance improvement is measured, not as good as the WDD result.
  • "Defrag+Auto" refers to the read time after enabling both the "Optimize Files" and "Optimize Boot" function in Windows, and then running JKD. An improvement of 11.5% is recorded, still slower than WDD.
  • "Defrag+BartPE" refers to the read time after rebooting into the BartPE environment (running from a CD) and then running JKD from the command line. The result was almost as bad as no defrag at all, and I have no idea why. I tried this because the "Optimize Files" option seemed to be contradicting the file placement method of JKD.
  • "XP+Auto" is the result obtained after enabling both "Optimize Files" and "Optimize Boot" functions in Windows, and then running WDD, i.e. the best that Windows XP can manage.
  • "Basic Office" refers to the read time of all 802 test files, where no defragmentation has been done whatsoever, after the installation of Microsoft Office 2007 Professional (Trial).
  • "Full Defrag" refers to the read time of all 802 test files after several reboots and two passes using the JKD program, with no "Optimize Files" or "Optimize Boot" options enabled, using the default defrag option. A 21.8% performance improvement has been measured, still 8.5% short of the "Office Auto" below.
  • "Full+BartPE" refers to the read time after rebooting into the BartPE environment (running from a CD) and then running JKD from the command line. This performance improvement of 18.9% is 12.4% slower than "Office Auto".
  • "Office Auto" is the result obtained by allowing Windows XP to do its own defrag after enabling both "Optimize Files" and "Optimize Boot" functions.
The performance improvement ranges between 11% and 21%, with slower results obtained when both built-in Windows functions are employed; "Optimize Boot" is enabled by default and does not appear to cause any performance loss, only "Optimize Files".
An average 16.8% performance improvement over no defrag at all is pretty good, better than the freeware Auslogics Disk Defrag program, and not too far behind the "heavy hitter" commercial packages. This is more remarkable because JkDefrag does not do any specific processing of the files listed in the "layout.ini" file.
This image is how the drive looked before the first defrag test was done. The yellow blocks are fragmented files, the pink part is the MFT reserved space, darker green files are "spacehogs" and lighter green are "normal" files. When the defrag has finished (as in the top image) the directories are stored first (bottom left), then there is a 1% free space gap, then the "normal" files, then another 1% gap, then the "spacehogs".
The file at the top of the top image is the pagefile.sys file, which was moved there during the BartPE defrag. In the bottom picture it is halfway between the pink section and the bottom of the image.
The next program to be tested is UltimateDefrag, from DiskTrix.

No comments: