Hello Roy

An update on my situation : everything is running fine since my last install. I haven't tweaked anything, I don't want to run into problems, so i'll leave it as is... as long as I don't touch those darn system cache settings, everything is A-OK...

Glad to hear Hugues,I`m sure it will work fine that way and by the way did you go NTFS of fat32,just curious? Have a good day and enjoy your new computer.

If any failure be it Memory/HDD/IDE controller occurs your liable to get this problem, Ive noticed in my experince that if you have a bad memory module altho the system may apear to work, when registry modifications are made to mission critical areas eg: HDD DMA settings .. and they are written corruptly next you reboot your sure to know all about it.

and just to put this thing to bed

I have seen this error on the following drivers

twice on WDC 120JB two different drives brand new.
one on a Maxtor 40GB 5400rpm ATA133 (dont know model)
twice on different Maxtor 60GB 7200rpm ATA133

etc.

I don't think it has anything to do with hard drives and I don't think it has anything to do with bad memory - since with the NVidia card this system is 100% reliable - its never actually ever crashed (I'm not kidding). I'd trust my life to this WINDOWS machine - and I think that says a lot personally.

I suspect that the ATI drivers are written outside of the AGP/PCI spec to improve performance - this isn't uncommon - one of the very first tricks ever used was invented by Matrox (and used by everyone to this day) for PCI/ISA video cards where the card doesn't lock the bus properly after a transaction to memory - since only one device can use a bus at a time.

Well I have seen the error on Both Intel and Amd Chipsets, I have seen this error on my VIA KT266A rig at home, which is just like yours 100% stable and has only ever givern me that problem once after installing on a 60GB maxtor drive.. couldnt recreate it. My Home rig has a ASUS Geforce4 TI4200 in it, The Intel systems i spoke of (I7205/I850E) have R9700 Pros..

I wasnt assigning the blaim PURELY to a bad memory module, but if you think "Its not a real cause" we can hook up and I'll prove it in a system with 1.0GB of ram which runs stable in every aspect untill you start using DMA on drives... I can just about gaureentee i can recreate it just buy just friging arround with the registry prior to installation with a few 3rd party apps..

Amd Athlon XP 1800+ 1.53GHz
Epox 8KHAL+ VIA KT266A
256MB Crucial PC2100 DDR
Asus 128MB Geforce4 TI4200
Creative Labs PCI
Maxtor 60GB ATA133 7200 rpm

And BTW the "third party registry tinkering" dosnt affect it at all with just 512mb ? .. so I gues bad memory isnt a cause by any stretch of the imagination sheesh.

And further to that what ever the true cause of this error is .. its very suttle wether your machine is doing it for reasons of bad memory integrity or poor ide cabeling, some crazy ass windows bug (which i doubt as the hardware affected seems too diverse), what ever.. it dosnt just always happen straight off.

I don't think anyone really knows whats happening here - thats why we post. I was just stating the known facts. I install an ATI driver machine turns to crap (this same card runs fine in my room-mates machine) - I remove it and go back to my old Geforce 2 and machine runs stable - and I don't know why (and after searching around google I don't think anyone really does). Doing some quick searches around google I've found there's a definate relationship between this error, KT266 chipsets and the ATI Raedon 8500 video card despite the motherboard, amount of ram or hdd used. I mean one guy I ran into fixed it by removing a usb 2.0 pci card...

I don't think its expected behaivor - so its either a known issue (bug) or something that can be fixed.

Anyhow - I've got an ASUS A7V266E with 1024 megs of PC 2100 ram - its not name brand (or if it is I can't remember what brand), but I can run memtest86 http://www.memtest86.com/ on it overnight with zero read/writer errors - and like I said this machine is 100% stable - its never crashed - even when it was acting goofy it just wouldn't let me log in, but I never saw a blue screen.

yea this is bit of madness

I mean one guy I ran into fixed it by removing a usb 2.0 pci card...- LOL wth

I have been experiencing the mysterious and elusive "Delayed Write Failure" for the past several weeks.

I have two animation studios which I travel between, and I carry with me an external ADS USB 2.0 enclosure with an 80Gb drive.

I have been experiencing the error on only this drive on both systems I connect it to.

If I stick the drive into the internal IDE inside the computer, the drive works fine, but I lose my fast transfer speeds.

I have tried two different ADS enclosures, about 5 different drives, between Maxtor and WD, in sizes of 40Gb, 80Gb, and 120Gb
The DAMN errors keep coming back.

As I'm reading this thread, and the common talk about the Radeon drivers, I have just installed a Radeon dual head video card on one of the PCs, and can't remember if the problems started before that or not.
But not sure if that would explain why it's on the second PC, which has a GForce2 GTS card.

I just installed a brand new 120Gb Maxtor drive, started copying data, and got the error.
I stuck the drive into the secondary IDE by itself, and successfully transferred 47Gb of data (it took over 10 hours) without error.
As soon as I put the drive back into the USB2.0 enclosure, the errors return.
I now installed the drive into my external Firewire enclosure, and so far no errors.
It acts like it's a defective USB 2.0 enclosure, but it would be quite a coincidence that two out of the three would be bad.

Could this just be a Window bug on Win2k and XP.

I'm amazed how loooong this thread has gotten ! Way back in the beginning I had this problem when I tried to install newer drivers for my ATI Radeon Card. The old ones worked fine. I managed a work around by installing the drivers manually. If I used the install program I got the delayed write failure but not if I unpacked the drivers and then added them manually myself. Well, recently I tried to install Unreal2, which wouldn't play with the drivers I had installed and so I downloaded the latest catalyst drivers from ATI and no way could I get them installed without the famous delayed write failure. Even trying to install manually, which worked before didn't make it this time. In the end I bought a new computer with a GForce card. I needed an upgrade anyway.

Huges how is it going with your system and the problem you were having ??

I actually gave up on this - switched back to a NVidia card. ATI cards seem to have lots of driver issues actually - check this out,

http://www.cs.yorku.ca/~dfiguero/tech/faq.html - while doing tech support for pdfmaker I found that while desk95.exe was running pdfmaker.dot (a macro for making pdf files using distiller) would cause word to run really slowly.

Kill desk95.exe off - everyone was happy...

Yeah the problem certainly doesn't have to do with the Hard Drive itself. I am experiencing the same problem and I have a 60 Gig WD hard drive. However to everyone's dismay I do have a AIW Radeon 8500 w/128MB! I have heard that Radeon's drivers are by far the most corrupt. They cause many problems on your system. I have gone through at least 100 reformats since i got this card. I defintely was unable to use the second update Catalyst driver but once the third update, most recent to this posting, I was able to use the card to its full functionality.
My problem started when I woke up one morning because for some reason the sound of my computer was getting on my nerves so I shut down my copmuter. I went to class that morning and when I got back during my break I discovered my problem. At first it took the coputer about 20 minutes before it displayed a promt saying that it had to use a back-up of the registry. Then after I restarted it, it then said Delayed Write Failure, multiple times. After restarting and doing some registry repairing, I still get the same message, even with a new regisrty installed. (You can input a new registry through recovery console. FOr information on how to do this, E-mail me.) But maybe my hard disk did fail. I use my computer for tunnel hosting for my X-box. I also do a lot of video capture, and if you know anything abotu video capture its vary tedious, and very hard on your drive. (Most Capture drives cost around $500). I'm not spending that, so I rty to on a regular HD. However the night before the incident, I was watching on movie, using Dual Desktop and left the movie on falling asleep. Of course there would be no standy by or suspension of hard drive. Now I haven't had this problem before. Only until I installed the drivers. And that would be the first night that I watched a movie on my computer. Now here is a list of components installed:
60 Gig WD HD
1.0 Ghz Athlon
2/DVD/CD-R/RW Drives
AIW Radeon 8500
Sound Blaster Live! Value
Dual Monitors
Two NIC Cards, running a 43Mbit Bridged connection (MAC Bridged T3 connection, full-duplex)

I had a problem with inetrent, intranet, and network before and it turned out to be Qos (Quality of Service Scheduler)

But that was no way related to this problem. My solution to this instead of reformatting. Purchase another drive. Make it master and slave your other drive. You don't lose your informationa dn you always have a back-up drive. The second solution is Norton Ghost and third and by far most reasonable is Tape Back-up. Tape-Back is the most secure and easy way to back up your computer.
Now I'll wait for replies and see what everyone has to say. I need to fix my computer, being at a tech school with a broken computer is an extreme handicap!

Hello folks
Just for your information: I'm having the same problem with the "Delayed write failed" since I installed windows 2000. My pc has on the primary channel a samsung drive configured as master, a maxtor drive as slave, old ATI Rage II and I'm experiencing the problem just with the maxtor. By reading all the entries that I found on google about this problem I'm quite sure that it occurs just on systems with maxtro/WD drives and ATI cards...

Hey.
Just to followup on my problems with this "delayed Write failure issue".

I found in my instance, that the problem was due to the wrong drivers being used for my external USB drives.

I think the reason the delayed write failure error was popping up was because it was a legitimate error.
As the drivers to my USB drives was causing my drives to do very strange things, from cycling on/off to locking up the computer.

So, for me at least, it had nothing to do with ATI drivers, as I'm using a GeForce 2 card. But it definately was a driver issue.

Brian
---------

Sapphire ATI 8500 128 (has anyone had these failures on the 64mb version? The stock ATI non-OEM version [the OEM version is made by Sapphire]? ) corrupts a Maxtor on a standard Supermicro Intel I815 board. Cache settings were altered (a long, long time ago), and system was stable. System is happy as a clam when flipping back to the internal graphics card. Can use the 8500 flawlessly so long as you don't reboot after installing the driver. The reboot, of course, leads to a cascade of "delayed write failed" errors.

Manual driver installation doesn't help. Have tried Catalyst 3.0 and 3.2, will try 3.1 next.

See you in a reboot.

P.S. Anyone for a class-action suit?

I could never install any Catalyst driver 3.0 and up without screwing up any of my maxtor drives. That computer is now my second and is working fine with earlier Radeon drivers. I guess the thing that still interests me is exactly what is happening. I can't understand why installing a certain iteration of graphics driver should screw up, windows, the hard drive, or whatever it is that gets screwed up. I wish some computer wiz could explain what generates this cascade of delayed write problems.

Rob

I disabled fast-user switching in MSConfig, switched system->advanced->Performance->Advanced->Memory Usage to best performance for Programs, and installed the Catalyst 3.1 drivers.

After 3 reboots, no delayed write failed errors. I'll try switching things around and finding whether one or another is responsible for the breaking.