One or more disks are failing error Ubuntu 10.04

Boot to liveCD and run a full fsck using the -a switch

You may also want to grab a copy of the ultimate boot cd which has the low level testing tools from the manufacturers to see if anything is amiss. The tool is free and should find any issues.

I spent half the day today testing the drive... the drive in question contains the PC OS, so it's NTFS. I used the chkdsk /r command from the XP console and it did detect errors and fix them. I also used the Ultimate Boot CD and used the SeaTools program for DOS. The short test failed for a SMART error, then the long test actually passed. Then I did another quick chkdsk, and it passed.

Then, I boot into LInux and after about 10 seconds of being into the desktop, a screen pops up saying that the drive is suffering from health issues, and the warning is "IMMINENT FAILURE". I go to the SMART readout and it says that there are 5116 sectors under reallocated sector count.

Suggestions? Is this drive failing, or is this a Linux issue like the prior version that everyone had trouble with? If it has bad sectors, is there a tool that I need to use to remap or bypass those sectors? And, should I just tell Linux to ignore the failure messages?

Thanks folks!

fsck checks file system. Here alert is on hardware, and is trigerred by a SMART daemon, which read disk autodiagnostic.
The SMART error seen in Windows XP could be the same reported by Ubuntu.

I installed smartmontools
sudo smartctl -a /dev/sda
give me for both disk
  5 Reallocated_Sector_Ct 0x0033 200 200 140 Pre-fail Always - 0
Unfortunetly I have a Western Digital and a Samsung.

Else smart daemon has problem with your model of hardisk
Else your disk has problem.

Could you try "sudo smartctl -a /dev/sda" ?

You could also use Windows software to confirm number "Reallocated_Sector_Ct".
If value is confirmed on XP, backup your disk.

Okay, I ran some more tests on the PC side and it came up negative for anything. Wow. Then, the same amount of sectors are bad on the LInux side. However, the drive in question is a physical drive setup for just XP, and the 2nd physical drive is setup for Linux.

I ran the above command line like you said, and I received the following:

SMART Self-test log structure revision number 1
Num Test_Description Status Remaining LifeTime(hours) LBA_of_first_error
# 1 Extended offline Completed: unknown failure 90% 1002 213567895
# 2 Short offline Completed: unknown failure 90% 1002 734283821
# 3 Short offline Completed: unknown failure 90% 999 680433291
# 4 Extended offline Completed: unknown failure 90% 999 172016879
# 5 Short offline Completed: unknown failure 90% 994 670149241
# 6 Short offline Interrupted (host reset) 10% 899 -
# 7 Short offline Completed without error 00% 211 -
# 8 Short offline Completed without error 00% 210 -
# 9 Short offline Completed: read failure 90% 200 977206
#10 Short offline Completed: read failure 90% 196 977206

  5 Reallocated_Sector_Ct 0x0033 001 001 036 Pre-fail Always FAILING_NOW 5116

So, Linux is being consistent, the PC is not. Hmmm...

Does it make a difference that this was a cloned drive from a former PC system? The old drive may have had bad sectors, as well. This current drive is over a year old, the prior one was about 4 years old.

Suggestions from here?

I'd like to get a 1Tb for the PC, and give the 500Gb we're talking about to Linux, but are there steps I should take to ensure that this is not a physical issue? Also, if I reformat the drive, or perhaps low-level format, will that clean this up?

Any further suggestions are very helpful. Thanks folks!

Also, again, I remember that when I originally installed Ubuntu from the Live CD, the original 9.10 without upgrades, I had the same problem. When I downloaded updates, that issue was gone. Then, when I upgraded to the 10.04, the issue is back. Now, I have not heard any strange sounds, there have been no read / write errors to my knowledge on the PC side of things, but it's just so odd that Linux is picking this up now.

The way disk controler manage bad sectors change from one disk to another, so cloning disk could be a problem.
You could have cloned table of bad sectors.
I'm surprised a 1 year old hard disk has problem, even if it is a 10000rpm.

So firstly, do a BACKUP.

I think that in your case, the best should be:
    low-level reformat to close possibility of cloning issue
    doing frequently backup in case you have

I typed enter by accident.

     And check regularly from both side (Windows/Linux) how evolve SMART indicators.
     If they diverge again post a bug.

I'm sorry. I don't see immediate solution.