Were you actually monitoring the S.M.A.R.T. information before you suspected that the drive was failing, though?
S.M.A.R.T. data is only meaningful when it's monitored over a period of time. The actual numbers are fairly meaningless, and are bound to vary from drive to drive, but if you suddenly find that, for example, spinning up the disk is taking longer than normal, then you might suspect that there is something wrong with the motor/spindle assembly.
S.M.A.R.T. is far from being perfect, but it is another tool that we have at our disposal. All modern discs support S.M.A.R.T. so you might as well enable it. If nothing else, you could replace the disk after, say, 10,000 power on hours, or 2500 spin up, (wash, rinse, spin), and spin down, cycles. I.E. Preventative maintainance.
> It should be trivial to just grep the kernel log for the error and mail it
> to some address.
Exactly, you can do this in user space, using tail -f and a perl script.
> Another way to prove a dead drive is (after a backup) to drop it some
> until it made a nice broken drive sound (a very high pitch shriek) and
> bring it to the store that sold it to you (ideally a small one as they are
> less technically challenged then big chains).
I think they would probably say that you voided the warranty.
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to email@example.com
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/