Disk Drive Issue

Matir · 01-28-2005, 07:40 AM

I have an ASUS P4C800-E Deluxe motherboard, with onboard RAID and SATA controllers. Last week, I had an apparent failure of my primary hard drive, a Seagate Barracuda (SATA) 120 GB. It started throwing i/o errors like crazy, and eventually even became noisy. So now I'm working on getting it replaced. This morning, I get up, and commands aren't working on my box. It tells me "bash: ls: command not found"... then a few commands give me "commandname: Input/Output error." All the while, I have a sinking feeling in my chest, wondering what has gone wrong now. I reboot the box and the hard drive is not even recognized by the BIOS. A couple of reboots later, I shut the box down fully, open it up and push on the IDE Connectors, though none feel lose. I reboot, and the hard drive works again. So, was this just a fluke incident? Or has something much worse died, such as my motherboard? It seems to be running fine for the moment, but I fear another such problem. And, if my SATA drive is good, I don't want the hassle of returning it for warranty. By the way, memtest86 has cleared my memory, if anyone thinks of that.

jailbait · 01-28-2005, 10:55 AM

"A couple of reboots later, I shut the box down fully, open it up and push on the IDE Connectors, though none feel lose. I reboot, and the hard drive works again."

I once had a computer in which the memory chips were not soldered in. The memory chips would work loose from vibration. Whenever I began to get memory errors (It usually took about 3 months.) I would push in all of my memory chips and the problem would go away. Maybe vibration loosened your hard drive cable and you fixed the problem by pushing it tight again.

------------------------
Steve Stites

Matir · 01-28-2005, 11:20 AM

I've just run SMART testing on hda and hdc... hdc reported a read error (bad sector, I'll deal with it later) and hda ran a LONG smart test with a clean bill of health. I guess I'm just paranoid after my hard drive failure last week (which, given the noise it began to make, i am fairly certain was a real failure). Any idea how, say, an IDE controller chipset failure would manifest itself? Could it affect both SATA and PATA?

Matir · 01-28-2005, 01:54 PM

I was looking through the logs of this 'event' and saw a number of messages such as:

Code:

Jan 27 15:17:13 [kernel] hda: status timeout: status=0x80 { Busy }
Jan 27 15:17:13 [kernel] hda: drive not ready for command
Jan 27 15:17:41 [kernel] ide0: reset: success

This was the only status message, no BadCRC or any of the other ones normally associated with a failing hard drive. And as I said, it passes SMART. Though, I'm not quite sure what some of the SMART fields mean and how to interpret some of smartctl's output. So anyone with any idea on these error messages? Bad cable, perhaps? IDE Controller? Transient error?