ATTENTION: You are viewing a page formatted for mobile devices; to view the full web page, click HERE.

Main Area and Open Discussion > General Software Discussion

Sysinternals PageDefrag: Good, Bad?

<< < (4/6) > >>

Steven Avery:
Hi Folks,

, I never saw any noticeable benefit from defraging, either using Windows' own or paging files PageDefrag or a few years back on my old machine when I owned a legal copy of Diskeeper -rgdot
--- End quote ---

 I do know that systems can fairly easily get to 25K or 50K of fragments.  (Yesterday I defragged a system with over 50K at work, and it is from a fairly lightly-used system of one of the managers, simply never had been defragged.)  Clearly that that causes a lot of extra disk access and also could complicate recovery procedures and possibly lead to data loss at times.  Thus even if  you don't get involved with any high-tech defrag, I think at the very least an occasional defrag lite (Auslogics, Defraggler from Piriform) in the background has to be a good idea. 


Midnight Rambler:
PageDefrag doesn't function in Vista.  After trying it on Vista SP1 a few times and finding that it didn't work, searched web for explanation and found forums stating so.  Also apparently no developer plans for future functionality.  Works fine on XP SP3, however.

Regarding Sysinternal utilities as a whole, there is a suite and a nice manager available that I'd recommend.

Judging by the last post in this thread it does work under Vista?

12GB ha!  XP (32bit) can only use 3GB of Ram and most of the laptops I have seen with Vista pre-loaded only come with 3GB as well.

As mentioned, JKDefrag automatically downloads and uses PageDefrag from SysInternals very well by itself.

I use JKDefrag in this manner: 

1.  I first delete all the temp files I can by using the last page of the GUI and optimize the registry as well.  CCleaner style. 
2.  I also turn on "show hidden folders/system folders" in Tools/Folder Options and delete all the folders that are in BLUE and start with $NT (Service pack uninstall), as well as empty the recycle bin again.  I've never had to uninstall an sp or hotfix.
3.  I then set XP for "No page file" and reboot. 
4.  Then I run JKDefrag GUI and move everything to the END of the disk. 
5.  I then reset the Page file to whatever I feel is optimal based on the computer, typically the standard 1.5x Ram, accounting for any shared video cards, and reboot again.  This places the page file as close to the outside edge of the disk it is able to find in a contiguous space.  I have used 2x Ram sometimes when the computer has more HDD space than ram.
6.  I then re-run JKDefrag GUI in "last access" mode making sure I specify any additional space hogs such as the c:\windows\installer\* directory. 
7.  From that point forward you can leave JKDefrag Screen Saver in whatever mode you like best and set it to wait an Hour before it kicks in, and then only defrag if the last defrag was more than 24hours ago.

Works like a charm!

An interesting thing to note is if you have the time to sit there and watch JKDefrag, you can learn all sorts of things about your computer and usually run across several other large TEMP folders you never knew you had that can be deleted as well, such as MSO, various folders in the "all users" portion of Docs and Settings, and any folder with a hexidecimal name,   Just stop the GUI, wait for the screen to reappear, go into explorer (with hidden/system turned on) and delete that folder, empty the recycle bin, then restart JKDefrag.

Oh, one more quick note about older computers with not enough ram or DIMM slots to upgrade, they sell compact flash to IDE adapters for $5 on Ebay that you can buy and put a 1GB+ compact flash card in, then put that on your secondary IDE channel as slave below your CDROM then once formatted you could completely move your Page file to that disk and even though it is not going to be as fast as real RAM, it is definitely faster than a page file on your primary IDE mechanical drive since disk I/O operations per IDE channel is the hindrance.


[0] Message Index

[#] Next page

[*] Previous page

Go to full version