Hardware Analysis Community Forums http://www.hardwareanalysis.com/content/topic/2458/ 2009-11-24T23:25:22-05:00 http://www.hardwareanalysis.com/content/topic/2458/ /images/halogo.gif text/html 2007-11-27T15:43:29-05:00 Michel Merlin http://www.hardwareanalysis.com/content/topic/2458/?o=160#520619 I see that this annoying problem has existed for years, with no solution. Thx to anyone who could help however...<br /> <br /> I have had the same problem, with little differences, for months now.<br /> <br /> The system (an old desktop used mostly to make backups of other systems - and finally for other things like TV) is Gigabyte GA-7VA Rev 1.x (chipset VIA KT400), Athlon XP2000+, Gigabyte GV-N66256DP (nVidia GeForce 6600), three HD racks, Maxtor ATA 133 with Promise chip additional HD controller PCI card; using mostly 250GB 3.5&quot; Hitachi HDs (up to 3 at a time), and 2.5&quot; HDs (one at a time).<br /> <br /> That system has worked with no noticeable problem for backups using different programs, including TIH (Acronis True Image Home) v9; the problem appeared immediately when I &quot;upgraded&quot; to TIH_10. See details in the thread before and after <a class="ext" href="/action/r/http://www.wilderssecurity.com/showthread.php?t=185256&amp;page=6#post1120843" target="_blank">http://www.wilderssecurity.com/showthread.php?t=185256&amp;pag...ost1120843</a> &quot;TIH_10 corrupted partitions when only READING them&quot;.<br /> <br /> My last incidents (not reported yet) are all the same: I launch TIH_11, I do nothing in it, I just leave it open, and after one hour or two, the &quot;Delayed Write Failed&quot; pops. WIth no TIH open, the error never comes (even in years of heavy operations).<br /> <br /> So it may be indeed that (as interestingly proposed by cow pits on Sat 3 Sep 2005 02:02 BST) this is a &quot;generic error&quot; in Windows that MS remains unable to figure after years, happening mostly during large flows of data (like copying 1GB), mostly on VIA chips, and mostly under a few risky applications or even versions (TIH_10 and TIH_11 are almost unable to work for me, but TIH_9 was almost without problem).<br /> <br /> Many people have been hit with TIH but unfortunately Acronis (like Microsoft or a couple other big corps) seems allocating more forces to PR than to curing the issue; so, Thx to anyone who could help.<br /> <br /> Versailles, Tue 27 Nov 2007 16:43:45 +0100 text/html 2005-09-12T12:14:23-05:00 Matt B http://www.hardwareanalysis.com/content/topic/2458/?o=140#322738 The problem for me is that I`m connecting this external drive to my laptop, so putting a new PCI FW/USB controller in isnt really an option.<br /> <br /> Sure, I could buy a PCMCIA controller, but I`d rather not have to spend the £30 or so.<br /> <br /> As I mentioned I was going to, I attempted to copy the data under linux, but sadly the live DVD I tried only had support for reading NTFS, not writing too.<br /> <br /> I`m pinning my hopes on being able to get my hands on a copy of the MS hotfix for a problem reported in their Knowledge Base which has similar symptoms.<br /> <br /> Until then, my internal drive gets more and more full !<br /> <br /> Matt text/html 2005-09-10T15:17:01-05:00 Jeff Monteith http://www.hardwareanalysis.com/content/topic/2458/?o=140#322074 Well, I'll give anything a try. I just ordered the following USB2 controller card for $10.99..... let's see what happens. (from here <a class="ext" href="/action/r/http://www.xpcgear.com/usb20pcihosc.html" target="_blank">http://www.xpcgear.com/usb20pcihosc.html</a>)<br /> <br /> NEC Chipset USB 2.0 5-Port (4+1) PCI Controller Card for Desktop PC (Retail) (FREE GROUND SHIPPING) Features ( USB 2.0 PCI HOST CARD) <br /> <br /> Add fast USB2.0 connectivity to your computer with this 5 port USB2.0 PCI Controller Card. It supports digital cameras, printers, scanners, keyboards, mouses, speakers, modems, USB hubs, game pads, MP3 players, USB readers, PDAs, plotters, removal storage bays (HDD, CDROM, CDRW)... etc. <br /> Best performance NEC host controller chip <br /> Compliant with USB Specification Revision 2.0 <br /> Data transfer rate can be 1.5/12/480Mbps <br /> Compliant with OHCI Specification for USB Rev 1.0a <br /> Five downstream port can handle high-speed (480Mbps), full-speed (12Mbps) and low-speed (1.5Mbps) <br /> 32-bit 33Mhz host interface compliant to PCI Specification release 2.2 <br /> Support PCI-Bus Power Management Interface Specification release1.1 <br /> Support up to 127 devices <br /> 5 Port (4 external ports plus one internal port) <br /> Support devices hot swap and wake-up <br /> Support PC Windows 98SE / ME/ 2000/ XP text/html 2005-09-10T07:22:39-05:00 NT Doc http://www.hardwareanalysis.com/content/topic/2458/?o=140#321985 I'm telling you as I said months ago. Take a new USB 2 controller card that is using the NEC chipset and stick it in your system and give it a try. I have a $3,000 <a target="_blank" href="/action/r/http://www.alienware.com/">Alienware</a> computer that had the same issues. I had a system I built myself that had the same issues. I took a USB 2 controller card that was using the NEC chipset and it fixed the problem on both comptuers. I now run a 250GB external drive in 2 units. An IOGEAR with a Firewire 800 and USB2 and an IOGEAR with only USB2. They both now work great on either computer. I can switch them to either computer, use USB2 or Firewire and no delays or hangs ever.<br /> <br /> I have 2 friends at work that also used a controller with an NEC chipset and it fixed the problem for them too. One had a <a target="_blank" href="/action/r/http://www.dell.com/">Dell</a> and the other had an home built unit.<br /> <br /> What the heck for $20 - $30 bucks if it doesn't work, take the card back and get a refund. Well worth the trial I think.<br /> <br /> Let us know how it goes for you. text/html 2005-09-09T14:08:00-05:00 cow pits http://www.hardwareanalysis.com/content/topic/2458/?o=140#321752 good luck. I highly doubt it's a hardware issue. i'm sure you'll have no problems with linux. <br /> let us know... i'm curious.<br /> <br /> it's the same crap with windows support for large ram, bogus if you ask me that you'd have to add a switch in the boot ini for xp so it recognizes more than 2Gig of ram. oh well... text/html 2005-09-08T12:06:57-05:00 Matt B http://www.hardwareanalysis.com/content/topic/2458/?o=140#321409 Hrm.<br /> <br /> I have received my new enclosure (which looks somewhat sexier than the last one), but sadly it doesnt cure the errors.<br /> <br /> I`m going to attempt to copy the data in linux, which should give some indication as to whether the problem is a Windows one, or is truly hardware-based.<br /> <br /> Wish me luck ...<br /> <br /> text/html 2005-09-04T14:33:45-05:00 cow pits http://www.hardwareanalysis.com/content/topic/2458/?o=140#320086 i had an issue with my 120G sata internal drive until it finally died and i haven't replaced it. <br /> i think windows just has issues with anything larger than 120 and it doesn't know how to manage the buffers, at least that's what i think. <br /> i would at this point recommend partitioning the drive with at least two partitions this way you'll know if it's a large drive issue or even a transfer rate issue. text/html 2005-09-04T13:53:16-05:00 Jeff Monteith http://www.hardwareanalysis.com/content/topic/2458/?o=140#320069 I have given up on this issue. I eventually took my 250GB external drive and placed it inside my computer, and then placed an 80GB drive in the external enclosure. I haven't seen this error since. <br /> <br /> In my experience I received this error every time I placed a drive greater than 100GB in the external closure. As a result, I just bought a 2nd enclosure and placed another 80GB drive in it. Not ideal, but no more errors.<br /> <br /> text/html 2005-09-04T09:09:44-05:00 Matt B http://www.hardwareanalysis.com/content/topic/2458/?o=140#320033 Cow,<br /> <br /> Thanks for the reply.<br /> <br /> Yeah, its running as an external drive. The only other peripheral I have plugged in is a USB mouse, standard wired variety.<br /> <br /> By the time I first used the drive, I had already installed XP SP2, so I dont believe I got the &quot;High speed USB device plugged into a slow speed port&quot; message ever came up.<br /> <br /> I find it confusing that the drive would work without a problem for 3 months or so, and then suddenly start producing these error messages.<br /> <br /> I have since bought a new enclosure, with a different IDE -&gt; USB bridge chipset, and am hoping that this will work. The old enclosure has a Prolific chipset.<br /> <br /> Heres hoping ! text/html 2005-09-03T01:02:25-05:00 cow pits http://www.hardwareanalysis.com/content/topic/2458/?o=140#319689 yes this problem does exist but i haven't had one in a while.<br /> mine was a with an SATA drive internal. however, from what i can tell about yours is that it's external, correct?<br /> may i ask what kind of chipset you're using? when you plugged in that drive for the first time, did it ever display a yellow balloon after the install that &quot;this device can perform faster&quot;? do you have a wireless mouse? <br /> the error is a generic one to where the OS is writing to disc and something interrupts the processor, hence it allocates the IRQ that is being used for the external disk. the OS at that point throws the error as it didn't finish. i'd recommend to try removing all your external peripherals, printers and such just unplug them all and run tests on your drive to see what happens. then plug them again and do the same tests. <br /> it's frustrating as it's a generic error.... good luck<br /> text/html 2005-09-02T22:08:32-05:00 Matt B http://www.hardwareanalysis.com/content/topic/2458/?o=140#319635 I, too, am having this problem. And I, too, 9 months after the last post to this thread, am unable to find a solution to it.<br /> <br /> I have an IcyBox (<a class="ext" href="/action/r/http://www.icybox.de" target="_blank">http://www.icybox.de</a>) external USB2 / Firewire enclosure, and a Seagate Barracuda 250gb drive. It was running on USB fine up until the other day (some three months of use), at which point I began receiving consistent and reproducible &quot;Delayed write failure&quot; errors.<br /> <br /> The errors occur each and every time I attempt to copy a large file (I dont know exactly what the threshold is, but I can copy an individual files ~5mb and a series of such files, but not a file ~700mb).<br /> <br /> I bought a good quality, shielded FireWire cable and tried that. With that, I can copy the files, but very slowly, and the transfer rate is very irregular. So, all in all, not an ideal solution.<br /> <br /> I have tried a great deal of &quot;solutions&quot; over the past couple of days, and none have worked. I have disabled fast user switching, enabled &amp; disabled write caching, created/deleted a swapfile on the external drive, updated the firmware in the enclosure chipset, soaked my testicles in vinegar and coated them with icing sugar. All to no avail.<br /> <br /> Has anyone found any solution to this problem, or at least an official acknowledgement that the problem exists ?<br /> <br /> Regards,<br /> <br /> Matt text/html 2004-12-07T20:01:09-05:00 Andrew Koluch http://www.hardwareanalysis.com/content/topic/2458/?o=140#211889 Check out the Maxtor Website. Maxtor says that this drive has a &quot;feature&quot; called Write Verify and that they ship it with it ON! The good news, the default is that it will turn itself off after 10 power cycles. Which sounds like after several failed installations of XP. They have a utility called WVSET to turn it off with a boot floppy. If you are having copy problems with new drives, try either power cycling the drive 10 times or run the WVSET OFF command from boot floppy. text/html 2004-10-12T07:47:19-05:00 afirstname alastname http://www.hardwareanalysis.com/content/topic/2458/?o=140#193107 An update: Recover4All was only seeing legitimately erased files, whereas my files have just gone missing. I have since switched to GetDataBack for NTFS, which seems more suited to the task at hand, and hopefully it fares better. text/html 2004-10-12T04:31:00-05:00 afirstname alastname http://www.hardwareanalysis.com/content/topic/2458/?o=140#193059 I got hit with a Delayed Write Failure error today. Now my 250gb NTFS drive only shows ONE 650mb folder, and everything else is missing.<br /> <br /> I stopped using the drive as soon as this happened.<br /> <br /> Any tips for recovering the lost data? I'm running Recover4all, which sees 23,000 deleted files but so far indicates &quot;100% overwritten&quot; for most of them. Is that possible? Would my drive have been sitting there wiping itself out during the write failure?<br /> <br /> Help! I need this data back! text/html 2004-09-05T22:11:51-05:00 Pesche Helfer http://www.hardwareanalysis.com/content/topic/2458/?o=140#178883 Here is what worked for me:<br /> <br /> Ok, got this problem solved, too. Luckily, just before trying to reinstall XP, I found out that this problem is called the &quot;delayed write failure&quot;-bug on English OSs. Although it is a highly common XP bug, it has never been adressed by MS. I screened the internet for it and found several suggestions, which I tried all at the same time. I know, it's not the prober way, but I didn't wan't to go through &quot;install no image -&gt; try method A -&gt; fails -&gt; reinstall image -&gt; try method B -&gt; fails ...&quot;. So everything works fine again, without reinstalling XP. Here is what worked for me:<br /> <br /> - disable fast user switch compability in MSCONFIG<br /> - optimize memory usage for programs, NOT system cache!!<br /> - inactivate secondary graphic board (PCI)<br /> - inactivate the any Firewire controller on the motherboard in device manager (1394 host controller)<br /> - when installing new catalyst drivers, make sure to inactivate &quot;Fast Writes&quot; and &quot;VPU recover&quot; BEFORE restarting the machine! (new control panel is already functional!)<br /> <br /> regards<br /> <br /> - Pesche<br /> <br /> P.S. I simply tried all simple suggestions I found in the internet at the same time, not one by one. It's possible that only one or few of them might do the trick!