All, Using hdparm as a benchmarking tool is fun, but it is NOT representative of most real life workloads, so don't get too focused on it. In normal use a disk drive does some read/write activity, then seeks to somewhere else, and repeat. DIsk seeks take real world time measured in milliseconds. Ignoring special cases that do large chunks of linear read / write activity, the biggest thing you can do to improve disk performance is to have multiple disk drives performing overlapped seeks. That is why people running data centers build raid arrays with lots of disks, that way they can have each of those disks simultaneously seeking around. Particularly when running a database, it is a huge win. IIRC, the formula that HP uses for designing these big storage systems doesn't even take per disk linear throughput into account. The variables are: 1) Raid type 2) Number of spindles 3) rotation rate 4) seeks per second Greg -- Greg Freemyer Litigation Triage Solutions Specialist http://www.linkedin.com/in/gregfreemyer First 99 Days Litigation White Paper - http://www.norcrossgroup.com/forms/whitepapers/99%20Days%20whitepaper.pdf The Norcross Group The Intersection of Evidence & Technology http://www.norcrossgroup.com -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org