On 9/22/06, Adam Thornton <[EMAIL PROTECTED]> wrote:
I hope you're wrong about that sustained rate. Because 4 GB/hr is already slightly greater than 1 MB/s. Surely substantially better is possible.
The 1 MB/s that I used was conversion of the 4 GB/hr quoted. We know at some point in time you run into the maximum disk I/O throughput that Linux can achieve, but I meant to say even in a poor setup I would not expect that to be at 1 MB/s already. But there's always surprised (like the customer using 'scp' to test their network bandwidth). One of the surprises can be the burst-like flushing out dirty pages. Some algorithms and defaults still seem to be designed for desktops where you need to retain interactive response over high I/O throughput. Tuning disk I/O in a virtualized environment is not trivial, and general recommendations that work is even harder. Rob -- Rob van der Heij Velocity Software, Inc http://velocitysoftware.com/ ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
