Has the swap file size been altered? What the buiffer size on the HDD? How much "available" space is left on the HDD?
Main Topics
Browse All TopicsI have a user that is really getting slow performance from her 2 SATA harddrives. A speed check indicates 2.9 MB/s - whereas my own reaches between 50-70 MB/s (both without burst). Her PC used to be really fast, and the situation only started a couple of months ago. The PC has a 3.2 Intel cpu and 2GB RAM. No RAID is involved. I have a feeling it is a hardware issue since it was originally performing fine. The PC is roughly 2.5 years old. OS is Windows XP with all SPs. Drivers appear to be up-to-date. My question: what could cause this kind of degrading in performance?
This Question has been solved and asker verified All Experts Exchange premium technology solutions are available to subscription members.
Experts Exchange has been collecting answers to technology questions since 1996…3 million and counting! If you have a question, chances are we already have your answer.
If you can't find the exact answer you're looking for, ask our exclusive community of 50,000 experts. You’ll get a personalized answer from a trusted professional.
Thousands of free tech tips, tricks, how-to’s and tutorials are available in our peer reviewed articles section. See for yourself how smart our experts are, no login required.
Access the answers to your technology questions today.
30-day free trial. Register in 60 seconds.
Members of the expert community talk about why the experience at Experts Exchange is different than what you will find anywhere else.

Try it out and discover for yourself.
30-day free trial. Register in 60 seconds.
Join the community of experts here and help other tech pros by answering question in your area of expertise. You can earn FREE access to all Experts Exchange's premium features and resources.
Are these drives using IDE or SATA mode?
If IDE, check the IDE ATA/ATAPI controllers in Device Manager for the current mode. [Right-click on each channel; select Properties; then look on the Advanced Properties tab] You may have had a reversion to PIO mode. If so, note which OS you're using and I'll tell you how to resolve it.
garycase: drives set to SATA mode. As I mentioned, it was originally working fine. The OS was already mentioned in the original question: XP (forgot the PRO).
warlock: I understand swapfile and buffer - but why would this cause a slowdown on both drives?
loftyworm: I'll know more this weekend when I get a chance to open the housing and do some more checking / testing.
Look on the Advanced Properties tab for the SATA controller -- that should also show the mode.
As for your question ["... what could cause this kind of degrading in performance? "] ==> PIO mode transfers would cause exactly these symptoms.
.... sorry I missed the XP => I read that and simply forgot ... and then looked at the question zones and didn't see an OS.
How are you measuring speed ?
HDtune is good for this.
Also you might consider running the vendor diagnostics as well.
Most can be found on the free UBCD at www.ultimatebootcd.com
I hope this helps !
A trojan horse or key logger could regularly access the drive...resulting in a huge loss of performance.
Can you monitor the file access using FileMon ?
http://technet.microsoft.c
That allows you to see which application/process id accessing your files.
==> Look for an unknown process accessing the NTFS regularly
xema: couldn't be fragmented - it was new and empty.
bigschmuh: no trojan or key logger - the drive was new and the performance is the same from multiple machines over the network -and- from the locally attached workstation (which is scanned daily).
sysexpert: HD_Speed v1.5.3.64 from SteelBytes -- I have heard of HDtune, but I believe it would only show me the big difference between internal SATA drives and the external eSATA MyBook. I have read in forums that there can be problems using eSATA, but I didn't believe it would be that big! There must be something I'm missing.
Oh boy, that's what I get for having two related problems: slow performance! I apologize for my last two comments - they pertain mainly to my OTHER querey! Not here. Going back to THIS question:
xema: the pc will go online in my shop this weekend - I'll check what you suggested.
bigschmug: same as above
sysexpert: the info in my previous 2 statements apply.
garycase: the SATA controller settings should not have changed from what they originally were - when it was working just fine - but I will be checking that as well over the weekend.
r.e. the SATA controller settings changing ==> Windows automatically reverts to PIO mode access on an ATA device if it encounters ten successive errors on the device. If that happened, then it would have done that without any intervention on your part ... and without any error messages. Until XP SP2 this was a fairly common problem ... but with SP2 the "rule" was changed from ten errors to ten successive errors, and this drastically reduced the PIO reversions. It's MUCH more common on optical drives; but the same rule applies to hard drives ... and it does occasionally happen.
By the way, since you asked it here ["... what is the difference between SATA and eSATA? "] ... there is NO signalling difference. The performance should be identical ... same controller; same signalling; etc. The only difference is in the cables and connectors -- the cables are better shielded and the connectors are a bit sturdier.
garycase: it is definitely a PIO problem. Here are the results of several hours of research and testing:
Two SATA drives one WD 500GB (200 free) and one SAMSUNG 100GB (21 free)
Both drives indicate an average speed of 3.7 3.9 MB/s (w/o Burst)
Both drives are slow defragmenting one will not speed up the other one
3.2GHz CPU -- 2GB RAM -- XP Pro SP3 -- 43 Processes -- Virus Check OK
Adding ResetErrorCountersOnSucce
BIOS indates HDD capable of UDMA6 (set to AUTO)
Hardware Manager shows both channels running in PIO instead of DMA mode
Toggling between PIO and DMA, even with rebooting between, did not help
Hotfix http://support.microsoft.c
KB817472 would not install --- only works with SP1 --- SP2 already installed.
MasterIDCheckSum deleted and rebooted --- still in PIO mode
Uninstalling secondary driver and rebooting --- still in PIO
(did not try uninstalling master - - - I boot from there!)
Ran resetdma.vbs from http://winhlp.com/node/10 --- still PIO after reboot
Under HKLM\Hardware\Devicemap\SC
Activated BusMaster in BIOS --- still PIO
Ran again the VBS script and rebooted --- no change
In Device Manager, be sure the "Transfer Mode" is set to "DMA if Available" for all of the IDE channels. Note that changing this does NOT change the "Current Transfer Mode" (which is the problem). Be sure you do this for ALL of the channels listed under IDE/ATA ATAPI controllers.
Then go to the registry editor, navigate to the following keys:
HKEY_LOCAL_MACHINE\SYSTEM\
The last four digits will be 0000, 0001, 0002, 0003, and so on.
Under each key, delete all occurences of the following values (not every key will have these values, but delete them whenever they occur):
MasterIdDataChecksum
SlaveIdDataChecksum
Don't forget to check EVERY key (...0000, ...0001, ...0002, etc. => as many as you have).
Now reboot and see if that helps.
If not, go into the BIOS and change the disk access method to IDE. Then reboot and repeat the process above.
As I noted in my first comment, "... PIO mode transfers would cause exactly these symptoms. " ==> so I'm not at all surprised that's what the issue is. The key now is to isolate WHY the drives are in this mode. Are there other controllers shown in Device Manager (besides the IDE/ATA/ADAPI controllers)?
I did note you had said "MasterIDCheckSum deleted and rebooted" ==> and while I assumed that was just a typo (the actual key name is "MasterIdDataChecksum"), I wasn't sure you had confirmed the "DMA if Available" setting, so I suggested you do the whole process again ... and emphasized that you have to check EVERY key.
Is there not a BIOS setting for "SATA Mode" ?? Typically this would allow you to set "legacy" or "IDE" instead of "SATA". Not all BIOS's will provide for this, however.
One other thing ==> you indicated you hadn't deleted the Master driver in Device Manager & rebooted. Go ahead and try that -- the system will still boot :-) [That has no bearing on what will be seen at boot time -- and Windows will redetected and reinstall it during the boot process.]
Deleted the driver - rebooted - controller found again (new devices) with both HDDs "found" - rebooted again. Device manager still shows PIO. I'm beginning to suspect that the hardware itself is limiting the speed. I have no legacy option in BIOS. OR - I need to go back through the individual steps in sequence: delete the MasterIDDataChecksum and then immediately uninstall the driver before rebooting. What do you think?
A few random thoughts just to be sure something simple hasn't been overlooked here ...
"... Her PC used to be really fast ..." ==> I assume this means the SAME hard drives were in it and worked fine until this slowdown started happening ... right?
Post a copy of the Device Manager window ==> I'm curious if there are any "unusual" devices shown ... or anything I think you should explore further.
It wouldn't hurt (since you just did the delete/redetect with the controller) to repeat the registry mod I noted (deleting the MasterIDDataChecksum/Slave
Are you CERTAIN there are no "SATA mode" choices in the BIOS? You said "... The only options I have are non-RAID and RAID ..." ==> are there any submenus once you select either of those? ... and look carefully at the other areas of the BIOS to be sure it's not "buried" on another page.
Is the optical drive (or drives) using DMA? ... or are they also "stuck" on PIO?
A couple of experiments to confirm that this is an OS issue and not a hardware issue ...
(1) Boot to a Knoppix or Bart's PE CD and do a file copy between the drives (or from one of the drives to a USB drive) and see what kind of speed you get then.
(2) Do you have a SATA -> USB bridge device? If so, plug the drives into another system via USB and see if you get reasonable speeds when copying off of them. Don't modify anything -- you don't want to modify the drives -- but simply copying reasonably large file off of the drive will give you a good feel for the performance (or just run HDTune against it to see what kind of speed you're getting). In the absence of a USB bridge, you could simply plug the drives into another system to do this.
(3) Alternatively, replace one of the drives with a "known good" SATA drive and see if it also gets accessed via PIO.
Last steps in sequence:
Booted into OS
In Regedit, HKEY_LOCAL_MACHINE\SYSTEM\
0000 = Standard-Zweikanal-PCI-IDE
0001 = VIA Bus-Master-IDE-Controller (for DVD/Burner)
0002 = Primary IDE Channel (Microsoft)
Deleted values of MasterDeviceTimingMode, MasterIdDataCheckSum, SlaveDeviceTimingMode (SlaveIdDataCheckSum key not present - nothing connected)
0003 = Primary IDE Channel (Microsoft)
Device values not present as nothing is connected
0004 = Secondary IDE Channel (Microsoft)
Deleted values of MasterDeviceTimingMode, MasterIdDataCheckSum, SlaveDeviceTimingMode, and SlaveIdDataCheckSum
0005 = Secondary IDE Channel (Microsoft)
Deleted values of MasterDeviceTimingMode, MasterIdDataCheckSum, SlaveDeviceTimingMode (SlaveIdDataCheckSum key not present - nothing connected)
In the Hardware Manager, uninstalled Standard 2-Channel PCI-IDE-Controller
Deinstallation required reboot (system popup)
System showed New Hardware Found - Samsung, WD, and Standard 2-Channel Controller, plus Primary and Secondary Channels. The System then showed a popup and asked if the computer should be restarted NOW. I said NO. In the Hardware Manager:
Primary IDE Channel -- DMA set, but nothing showing as detected (probably the VIA)
Primary IDE Channel -- DMA set, but nothing showing as detected (probably SATA)
Secondary IDE Channel -- DMA set, Ultra-DMA-Modus 2 detected for both 0 and 1 (VIA)
(these are the DVD/Burner drives)
Secondary IDE Channel -- DMA set, PIO-Mode detected for device 0 (SATA)
Closed Hardware Manager and rebooted ---
Primary IDE Channel -- DMA set, but nothing showing as detected (VIA)
Primary IDE Channel -- DMA set, PIO-Mode detected for device 0 (SATA)
Secondary IDE Channel -- DMA set, Ultra-DMA-Modus 2 detected for both 0 and 1 (VIA)
Secondary IDE Channel -- DMA set, PIO-Mode detected for device 0 (SATA)
. . . . .
Checked BIOS (again): SATA Operation Mode = non-RAID&.choices are non-RAID and RAID
SATA1 = Hard Disk (Samsung --- 500GB)
SATA2 = Hard Disk (WD --- 120GB)
As an experiment, set the WD at UDMA 4 manually --- no positive results.
Checked all other settings regarding SATA or IDE --- nothing available
In the BIOS --- Loaded Optimal Defaults
Motherboard is AsRock P4V88+ P1.20 . . . . have NOT attempted BIOS update since it WAS working before.
. . . . .
Temporarily connected another SATA HDD in place of the WD and insured it was showing up in the BIOS. Except for the size (80GB Hitachi) all the parameters were the same.
New Hardware was found during the next boot. The HW Manager indicated the Hitachi was running in PIO Mode. Again deleted all master and slave DeviceTimingMode and MasterIdDataCheckSum values, uninstalled the standard IDE controller and rebooted.
New hardware found (again) and rebooted (again).
Result: Hitachi running in PIO Mode!
OK --- completely disconnected SATA drive 2, exchanged the SATA data cable on the remaining SATA drive 1 (Samsung --- OS) and rebooted.
PIO Mode again!
OK --- disconnected both users drives. Connected the Hitachi (previously detected as PIO), formatted, and installed basic XP Pro OS (SP1). Did not install any additional drivers other than those provided by Microsoft --- again PIO with the Hitachi drive!
Found on the AsRock site a specific SATA driver, downloaded and updated the standard IDE controller to the VIA driver offered. Now, under the HW Manager, there is only a single primary and secondary channel under Controllers!
Hooked everything back up the way it originally was. Updated the driver for the Standard IDE Controller using the VIA driver I had downloaded. Rebooted 2x. As expected, the last reboot was extremely fast. The Samsung 500GB running at an average of 80+ MB/s and the WD 120GB at around 60 MB/s and burst jumps over 100 (better than my own --- need to take a closer look at that).
Conclusion: the VIA driver had somehow degraded or become corrupted. XP decided to stepped in and exchange the VIA driver for a generic one (unseen, and without an error message of any kind), but it did not support UDMA under SATA for this controller, so sidestepped to PIO.
WE have got it solved! MANY thanks to garycase for staying with me on this one. Total time: 14+ hours (and that was only MY time) --- all because of a corrupted driver. Live and learn: don't go looking for a complicated answer to a simple problem. Thanks, Gary!
You're most welcome ==> as I noted in my first comment, the symptoms made it almost certain this was a PIO reversion issue. It was just a matter of isolating WHY it was happening. A strange result -- I'm surprised Windows didn't flag the device in Device Manager since it didn't have an appropriate drive ... but at least it's fixed !!
Business Accounts
Answer for Membership
by: loftywormPosted on 2009-06-18 at 11:30:14ID: 24660360
Does it effect both drives? I would suspect the FSB. I would get some diagnostic or burn in tools if I needed to narrow it down, but the simple is you will likely hav eto replace the MB, where the HD controller is.