> I question the utility of the hdparm performance. This doesn't appear to be random > reads.
Good point! Does anyone know how hdparm works? How accurate (in what sense) does the figure reflect the performance (esp data read-write access rate) of a tested hard drive?
