OK I will take a crack at this... The disk utility seems to be causing some panic and may or may not be incorrect depending on your specific hard drive. In reality the SMART implementation of the hard drive matters.
1) In disk utility note what the drive path is: as an example /dev/sda
2) install the smartmontools if they are not already installed
smartctl has a dependency on postfix so it will prompt you on how you want to configure mail. Unless you really want to configure mail properly I would choose none or local. (I have the smart tools email me notifications, but that is not something most people will need.)
sudo apt-get install smartmontools
3) from a command prompt and type
so in the example of the firdst drive (dev/sda)
sudo smartctl -A <path to drive>|grep Reallocated_Sector
Note: smartctl -A puts out a lot more information and you can look at it in its entirety by removing the grep. Since this is the particular attribute that is causing a stir I wanted to narrow the focus.
sudo smartctl -A /dev/sda |grep Reallocated_Sector
Now you should get output something like this
The only numbers to concern yourself with at the moment are the last one, the raw value, and the one immediately before "Pre-Fail", the threshold. The threshold will vary depending on the type of disk and manufacturer. Generally 36 seems to be a common number for desktop hardrives. For 2.5" (laptop) hard drives it varies more. An unscientific survey of the dirves in my laptops showed thresholds between 24 and 140.
5 Reallocated_Sector_Ct 0x0033 100 100 036 Pre-fail Always - 6
- if your raw value is a very high number (over 500) then it is likely not reporting SMART data correctly, get the disk utilities from the manufacturer (see below)
- If your raw value is over your threshold then you very likely have a problem and the drive should be replaced. run "sudo smartctl -A /dev/sda" and look at the output in its entirety. Additionally you should run the manufacturers disk utilities (see below). This will speed up the RMA process if it is still under warranty.
- If your raw value is over half of your threshold I would suggest keeping an eye on the drive at least. you might consider trying to RMA it but depending on the manufacturer it may not be "bad" yet.
- If your raw value is under half of your threshold then the drive is very likely fine.
Again this is personal experience, not concrete fact, but I have noticed a tendency for Seagate and Samsung Drives to report a few bad sectors on a regular basis. Western Digital however rarely reports bad sectors until the drive is starting to have problems. This may be just the luck of the draw for me or it may be a variation in the way the different manufacturers report SMART data.
MANUFACTURER SPECIFIC TESTS:
If you want additional assurance on the status of your drive I would suggest going to the manufacturers site and downloading their disk testing utility. Most will require you to boot into windows but some will have a bootable CD. In reality for the most part these disk utilities will basically be reading the SMART data, but since it is a program from the manufacturer I would argue that they are better equipped to determine the appropriate values for the data.
Here is a link to the disk diagnostic tool for Seagate:
This site (once you pick your specific drive) is for Western Digital
To put this in perspective I have multiple fairly large RAID5 arrays. Out of 5 1.5TB Seagate drives I have 3 with bad sectors(1, 3, and 6 bad sectors). Out of 5 1.5 TB Western Digitals I have none currently reporting bad sectors. I do not attibute the difference to WD being more reliable than Seagate but to a difference in the way the 2 manufacturers report SMART. Out of 12 other drives from various manufacturers and size only 2 are currently reporting bad sectors (14 and 9). None of these numbers are high enough to cause me alarm at this stage but I do have mail notification set to monitor the health of the drives so if/when one does approach PRE-FAIL I can replace it before it becomes a problem. Disk Utility is a great tool in that it puts SMART data in a visible element for the end user. But it has the unfortunate side effect of causing panic when the user is not used to seeing errors on their hard drives.
EDIT: I wanted to add a small editorial on Windows "check disk" chkdsk and other (non manufacturer) Windows disk utilities are pretty much worthless for determining the health of the hardware. What the Micorsoft tools check is whether the file system is intact, not whether the hardware on which the file system resides is intact. In most cases by the time chkdsk would detect a problem caused by hardware failure it is too late. It is for this reason that SMART was created in the first place. It is not fool proof, not even close (check the paper written by Google on the subject), but it does help.