Slowdown problems in Windows XP

DewFiend

Junior Member
Mar 17, 2002
1
0
0
I've recently been having some problems with Windows XP. I just did a clean install of XP around two months
ago, and it has been running perfectly until just the other day. It is slower to boot up and I get long pauses
when launching programs. When I get these pauses everything slows to a crawl, and the hard drive is
constantly being accessed. Once the program has loaded it speeds right along. As far as I can remember,
I didn't make any changes (although it seems obvious that something changed, so perhaps I did).
This is what I've done so far:

-- Virus scan (none found)
-- Checked Task Manager during pauses, which shows less than 5% cpu usage (so 95% is free)
-- Defraged Hard drive (twice)
-- Disabled any unnecessary services
-- Checked the event viewer for errors (none)

I'm at a loss, so any suggestions would be great. I REALLY do not want to rebuild, so leave that as a last resort.

My system is running at 1.2 GHz, 512 MB RAM, and the HD in question is a 60 gig 7200 rpm seagate barracuda IV

Thanks in advance!
 

OpalFrost98GT

Senior member
Aug 4, 2001
527
0
71
Try turning off system restore for a while and see if it clears up. If it does reenable it and see if the problem returns...
 

NogginBoink

Diamond Member
Feb 17, 2002
5,322
0
0
Are you getting low on disk space?

The best tool to diagnose performance problems is performance monitor. Without practice, however, it can be a difficult tool to use. I'd suggest you at leat poke around with perfmon and see if you can stumble onto anything obvious. And you'll learn lots as well even if you don't find the root cause of the performance problems.
 

MatthewF01

Senior member
Mar 1, 2002
728
0
71
Check your drive's cluster size, make sure its not at 512 bytes. If so, you might need partition magic to change it to [umm i forgot, but its something like 1.**kb]. I hear thats a problem...

I pulled the following from a Scot's Newsletter (www.scottfinnie.com)

NTFS PART III: HOW TO MAKE NTFS GO FASTER
> -------------------------------------------------------------------
>
> In the last two issues of this newsletter I've run one of my
> inadvertent series -- this one on Windows XP's NTFS file system. I
> had no idea I was about to touch off an avalanche of questions and
> reveal a large problem for Windows XP users. Or even better, that I
> was about to publish an NTFS performance solution.
>
> I explained about cluster sizes in NTFS in the first in the series:
>
> http://www.scotfinnie.com/newsletter/19.htm#filesys
>
> In short, Windows XP's NTFS file system is even more storage
> efficient than Win98's FAT32. And it can be at least as fast as
> FAT32. It's also more reliable. But there is one very big problem.
> When you install Windows XP as an upgrade of a previous version of
> Windows running FAT16 or FAT32 and convert to NTFS as part of setup
> or after the fact, in most cases you end up with tiny 512-byte
> cluster sizes. This occurs because of the way the data is aligned on
> the disk and the NTFS conversion process as carried out by
> Microsoft's Convert utility. PowerQuest's PartitionMagic 7.0 uses
> the Microsoft utility, so it has the same issues.
>
> The surprising truth is that some new PCs also arrive with 512-byte
> cluster sizes. So if you've got a new Windows XP box that runs slow,
> you should definitely check out what I'm about to explain.
>
> In the last issue, I alluded to a possible solution:
>
> http://www.scotfinnie.com/newsletter/20.htm#filesys
>
> A little-known program called Paragon Partition Manager, created by
> a group of Russian programmers working for Paragon Software, has in
> its latest version, 5.0, added the ability to dynamically adjust
> cluster sizes.
>
> Paragon Partition Manager 5.0:
> http://www.partition-manager.com/n_pm_main.htm
>
> Paragon Partition Manager 5.0 Features Details:
> http://www.partition-manager.com/n_pm_requir.htm
>
> Paragon Partition Manager isn't generally marketed in the U.S.,
> although you can purchase it on the Internet, where it sells for
> about $40:
>
> Where To Buy Paragon Partition Manager:
> http://www.partition-manager.com/n_pm_buy.htm
>
> None of the popular disk utilities marketed in the U.S. is capable
> of pulling off this feat yet. But Partition Manager does it, and
> does it well.
>
> To prove the point, I acquired a copy of Paragon Partition Manager
> 5.0 from the company and configured a test system. I had an existing
> drive containing a clean Windows 98 Second Edition installation on
> my trusty Compaq Armada 700 (the best Compaq product I've ever
> worked with) notebook PC. I ran a standard Windows XP upgrade
> installation, which took a while, but completed just fine. I also
> converted to NTFS. When all the files were copied and the changes
> made, it was immediately apparent to me that my performance eroded
> markedly. It took Windows forever to load, and disk-intensive tasks
> ran like molasses in January. In fact, I was surprised by how slow
> the machine became. I had been led to believe that 512-byte clusters
> slowed the machine down incrementally, but the reality was much
> worse.
>
> Next I used Windows' Disk Defragmenter to check the cluster size on
> my hard disk. To do that, you open Disk Defragmenter from Start >
> All Programs > Accessories > System Tools > Disk Defragmenter.
> Right-click the appropriate drive and choose Analyze. When the
> analysis is complete, click the View Report button. There you see a
> line that reads Cluster Size = XX KB. In my case, it showed 512-
> bytes, the smallest, slowest cluster size NTFS allows. The optimum
> size is 4K clusters.
>
> I was ready to try Paragon Partition Manager. I'll tell you upfront
> that there are two problems with this product. The first is the user
> interface, which needs help. But it's usable. The second is that
> before you make the cluster size change, block out several hours of
> time for your PC. Overnight might be a good idea. You may save
> yourself some time by running a defrag before you run the cluster-
> size conversion, but you'll find that Disk Defragmenter also runs
> very slowly when your cluster sizes are 512-bytes.
>
> The time factor thing is variable. While it took me four hours to
> convert the cluster size on a 12GB notebook drive, it took one SFNL
> reader only a few minutes and another one over six hours. I asked
> the Paragon people about that and they wouldn't commit to even a
> range of time you can expect this process to take. Reading between
> the lines, this large difference from PC to PC in the time it takes
> to run the conversion is normal.
>
> The user interface issue comes into play when you do the cluster
> conversion because nothing says "Convert to 4K Cluster Size." But I
> can save you that pain. Once you have Paragon Partition Manager
> running, select the NTFS drive whose clusters you want to convert.
> >From the program menu, choose Partition > Modify > Change cluster
> size. Dial the "Sectors/Cluster" spinner up to the number 8. Press
> OK. (If you select 4 in this scroll box, you'll get 2K clusters --
> not the desired outcome.) The conversion process requires that
> Windows XP reboot.
>
> Maybe you'll be lucky and have the fast-track conversion. If not, I
> can promise you this, it'll be worth the wait. As soon as the
> conversion completed for me my performance was back to FAT32 levels.
> All that was left to do was run Disk Defrag again, both to check the
> cluster size and also to defrag the disk. Do both things.
>
> A couple final notes on NTFS this week. First, I've received a ton
> of email about NTFS that I haven't had time to get to. Many offer
> interesting info or questions deserving response. I will continue to
> cover NTFS in future issues.
>
> The other point is that Microsoft is continuing to investigate
> issues people have had with slow NTFS performance on new Windows XP
> PCs. The company is working with some of SFNL's readers on that
> point. I hope to get some sort of report back from Microsoft -- and
> if so, I'll publish it in a future issue. My take though? The steps
> in this issue will probably fix your problem, assuming you're
> willing to shell out for the Paragon product.
>
> I'd like to thank SFNL readers Jerry Bass, Tom Synder, Tom Duda, and
> literally scores of others who volunteered excellent information,
> donated hours of their time, and asked smart questions. Many of the
> best things in this newsletter come not from me, but from its
> readers. And for that I'm eternally grateful.






maybe this helps you?