(Untitled)

[chronicfreetime]

I think I've straightened out the disk corruption issue. After considerable testing, I've determined that it only happens when using both SATA channels simultaneously, which means the controller (or its driver) is at fault. The disk controller is manufactured by Silicon Image, which is the new name of CMD Technologies-manufacturer of the old

Comments

[roadriverrail]

If you don't mind posting it, I'd be interested in your testing methodology.

[chronicfreetime]

dd if=/dev/zero bs=1024k count=1024 of=t; od t

Hmm, nothing. Maybe it matters what the bit pattern is? I write a C program which calls srandom, generates a large file of noise, then reads it back and compares. Still no errors. Maybe it only happens on some regions of the disk? Just to convince myself that I'm not hallucinating, I unpack SP2 twice and compare them. Still broken, and it even requires a chkdsk to fully clean up.

The SP2 archive is located on my boot drive, which is channel 0 on the SATA controller. On a hunch, I run my tester program on both drives at once. I start getting several errors per gigabyte, and even some on the boot drive. I try a few other things, to see if lots of interrupts or context switches alone is enough to trigger it, but it doesn't seem so. I start searching the web for a fix.

[scathedobsidian]

I've had two similar occurrences. One was a year and change ago when the controller on a motherboard failed, and the other was a week and a half ago, when the logic board of an HD failed.