In my case, the Lenovo Quick Test Disk Diagnostic Utility (had to be downloaded) revealed that there is an unreadable sector on the C drive, and that appears to be the cause of this error message from iaStor. It appears that since the drive can not read from the bad sector, the drive is timing out on read requests for that sector. The newly installed ShadowCraft Shadow Protect Desktop software would not create a backup of the C drive because of the bad sector. Image for Windows (still installed but just deprecated) however, has a setting that allows it to continue even if there are I/O errors. If necessary I hope to use that setting to get a snapshot of the drive that I can re-image to a new drive. Some app may need to be reinstalled if it's got a key file that includes the bad sector. But I'm counting on the rest of the image from Image for Windows being good to get the re-imaged computer up and running quickly for the client, with a new hard drive. The computer is a Lenovo ThinkCentre Desktop MT 7052, with 500G factory HDD.
I had this problem myself. I talked to Intel, GigaByte and Samsung with no luck. Then I found this. As Tim said. This is a most probably a power problem. Glitches in the SATA power cable or the SATA cable itself.
Seems Vista made the problem worse, see the "Possible issues with Windows Vista" link. I'm still running WXP and W2K3 and still having issues ranging from RAID re-mirroring to BSODs. Finally a solution from Intel. A new version of Intel Matrix Storage Manager that fixes this problem: 126.96.36.1993 (see the Intel Downloads link).
See the link to "Pauls Computer Service" for an article that may resolve your problem.
This problem has occurred a few times on my main home system a couple of times causing my RAID0 array to complain of corruption. This error would occur when a high-load power situation would arise (typically, when I would use the DVD drive in conjunction with all the HDDs), so I suspected a power issue. The rated wattage of the PS is well above the draw of the system so I reexamined the distribution of circuits and suspected I must have been drawing too much from a specific leg of the PS. I reconfigured all the power cords (adding another Molex cord -- it is a modular PS) and the problem disappeared.
I found this issue to be related to the driver version and not the hardware. I rolled back the Intel 82801G RAID driver to the version matching the BIOS and the problem has gone away. Consider rolling back drivers before attempting to swap hardware as it could save you a bunch of time and energy.
Details: I have a RAID1 setup w/ 2 WD2500JS drives on a GA-8I955X Royal motherboard (955X, ICH7R). It has been stable and error free since I built it in early 2006. I started having these errors over the past 3 months (usually 2 within 15 min), followed by an Event ID 5, "A parity error was detected on \Device\Ide\iaStor0". At this point the RAID array is degraded and needs to be re-mirrored. Once re-mirrored, the problem could occur again in hours or days. I realized after some time that the problem would occur while running SlingPlayer (Slingbox) on the machine. The errors appeared more likely when SlingPlayer was in full screen mode on the second monitor. Coincidentally, the audio would crackle when the problem occurred and continue until the player was reloaded. CPU (5-15%), paging (0-100/sec), disk I/O (virtually 0), network I/O (60 KB/sec) were all relatively low, much lower than the database activities routinely run on this machine without error. I read the adpu160m comments and went through the trouble of chasing the "bad hardware" notion. This is a process I wish on no one thus my contribution here. In the end, I had rotated in 2 other seasoned WD2500JS disks from a RAID0 array on a Sil3132 controller on the same machine. This included rotating SATA cables and trying the other 2 ports on the ICH7R. Nothing seemed to help and the "suspect" disks / cables performed without error on the Sil3132 RAID0 array. After discounting the hardware as the culprit, I noted some old Abit posts on F6 driver versions so I went at it from that angle. I have a tendency to keep drivers current. So the likely culprits were the ATI video, the Realtek HD audio and the ICH7R RAID drivers. If nothing worked there, I was prepared to rollback to an older Ghost image to factor out all those MS Updates and whatever else may have transpired. First, I rolled back the Realtek driver because I did not like the latest version anyway (ton of garbage language files in Temp dir). Nothing changed. Next, I concentrated on the RAID driver after watching DVDs and other video content streamed across the network produced no errors leading me to believe the ATI driver was not a likely problem. I rolled back the RAID driver from the latest 6.1 version to the 5.5 version in place during much of the stable period. The frequency of the errors was reduced but not eliminated. I happened to notice the BIOS version during POST was 188.8.131.522 so I took the chance and rolled back the driver to 184.108.40.2062 while keeping the 6.1 Storage Manager application. The problem essentially disappeared. I say "essentially" because I have had 1 Event ID 9 in the week since the rollback (SlingPlayer was running).
Thus my anecdotal conclusion: The Intel driver is the issue. Regardless of any issues with SlingPlayer, the Intel driver should be rock solid and that does not seem to be the case in this configuration. Matching the driver version to the BIOS version appears to be the best compromise.
See EventID 9 from source adpu160m for information on this event.