The first sign you’re likely to get of a failing hard drive is a SMART warning when you boot your computer. Something like this …
Which is exactly what I got recently on a relatively new 1TB drive.
SMART stands for Self-Monitoring, Analysis, and Reporting Technology and there’s really not much you can do except be grateful for the warning, get a replacement drive and start moving data off the failing one as soon as possible.
SMART records all sorts of information about your HDD, some of which is used to predict future drive fails. The most common is the Reallocated Sectors Count fault — which is what my dying drive has.
What that means is that the hard drive has found a read, write or verification error on a part of the disk, marked the dodgy sector as “reallocated” and moved the data to a working sector. As you can see above, this has happened 3,995 times so far. As a predictor of impending failure, it’s a good one.
The BIOS warning doesn’t give you much information about precisely what’s going wrong with your disk. To find that, you’ll need to add some software appropriate to your operating system.
The best package I’ve come across for Linux is simply called Disks. It’s included in Debian, Ubuntu, Fedora, Red Hat Enterprise Linux and CentOS, and typically filed under the Utilities menu. If it’s not already installed, just grab it with your usual package manager.
Of course, SMART can’t anticipate all hard drive failures. In the case of a sudden failure, your only realistic recourse is to your backups. If you don’t have any, and the data’s critical, you can get the drive rebuilt, but it’s going to cost in time and $$$s. A quick local Google found one firm with a turnaround of 5-10 working days at a cost of about $600. Considering the price of a new 1TB drive is only about $100, I’m happy to heed the warning!