Weird HD Progressive Meltdown

[kpep01]

This problem may have started earlier on an older (now gone) X86 box. The old box died what I thought to be a natural death of old-age - now I wonder.

Last November (2007) I plugged in all of the old hard drives to my AMD64 box, and copied the files over to my SATA drives. By May of 2008, the ADM 64 box was becoming unresponsive and locking up. Eventually, the core system files were wiped out.

The SD that I was using was less than one-year old, so, that should not have been an issue.

When the system eventually wouldn't do squat, I simply rebuilt using a spare SATA drive I had sitting around. I rebuild Gentoo on the new SATA, then copied my files from the old SATA onto the new one - it was a bit of a chore, considering the I/O errors, but, I got the job done.

After the file process copy was complete, I reformatted the old SATA, and copied my archives back onto that SD.

Again, this was last May (2008). The new SATA drive had never been used, so was clean.

Within the last week, the progressive SD meltdown began happening again. Recognizing what was happening, I was quick to ensure that my current config files/etc. were all copied to the older SATA drive, as well as all of the other important stuff on the newer drive.

This process was again complicated by numerous I/O errors, but, I did manage to save my files.

I then began the laborious task of rebuilding Gentoo on the newer drive starting with a complete reformatting of the drive (if there is something living on the drive, I wanted it gone.)

Now we come to the fun part.

I can't reformat the drive - whether using multiple partitions or using it as one single partition. There are numerous I/O errors, etc. This really shouldn't be the case on a drive that is less than 6-months old.

Most interesting among the error messages is (or seems to be in my paranoid mind):

[bunder]

Moved from Networking & Security to Kernel & Hardware.
_________________

[Janne Pikkarainen]

Bad cable, bad RAM, too much heat, or actually a dying drive.
_________________
Yes, I'm the man. Now it's your turn to decide if I meant "Yes, I'm the male." or "Yes, I am the Unix Manual Page.".

[manaka]

Could you post the output of dmesg to see the kernel messages?

The output of smartctl -a /dev/sda could also be of help to spot a dying drive.
_________________
Javier Miqueleiz

"Listen to your heart. It knows all things, because it came from the Soul of the World, and it will one day return there."

[pappy_mcfae]

A drive can fail at six hours old, six days old, six months old...for any number of reasons. When you consider the physics involved with surface mount chips, it's a wonder that most cutting edge electronic systems don't just fall into smoldering piles of ashes. When one considers the physics of how the information is physically stored on the drive's platters it's a miracle that data can survive at all.

If I were you, I'd take said drive and get it replaced. Don't worry about theories. If you still have the receipt, replace! I have a Maxtor that's going back soon. It went to sleep functional after I removed it from my tower. I went to do some funtoo experimentation, and the drive was dead in the water. It spins up, but no head accessing, clicking, or the like. It's less than two years old (5 year wty).

I have no idea what killed it. All I know is it was dead, and I was really glad that there was NOTHING of import on the drive. As NeddySeagon's sig says,

[aka.doode]

[cyrillic]

I would tend to agree with Janne Pikkarainen as far as the "too much heat" thing goes.

This is the number one reason I have seen my own harddrives fail.
Now, I make sure my computer cases have plenty of ventilation, especially around the harddrives.