Well, there is most likely no way to know before-hand that a disk will fail. But you could try to read the signs. And what a surprise, there is help already available. This S.M.A.R.T. thingy could be of help; SMART collects a lot of data of your disks, which can be used to evaluate how “well” your disk is doing. This could show some up-coming trouble.

To read the SMART-Data, there exists quite a few programs. For Linux you would most likely use smartmontool, and fortunatly, these tools are also available for Windows!

Getting and Installing smartmontool

1. Just download the latest Windows-Installer from sourceforge, which will install all needed binaries.
2. Edit the smartd.conf to meet your needs
3. Either run smartd in daemon-mode, or as a Windows Service

Configuration

The DEVICESCAN option allows to detect all available devices and add them to the monitoring list; but I like to tailor my config-files a little. So I comment-out the DEVICESCAN and add some options of my own:

   hda -a -o on -S on -I 194 -I 231 -I 9 -m sysmsgbox

This will monitor my 1st disk, and ignore some measures (-I 194 -I 231 -I 9) like thermal stuff. Besides that I would like to carry out all of the typical tests (-a), do selftests (-o on) and collect statical data (-S on). Finally I would like to get a popup (‘-m sysmsgbox’; this is just available in the windows version) if something is going wrong