Malcolm Baldridge wrote: > While it's handy for extra storage to "fill up" the master/slaves, the > performance is never good. It's especially bad because you have the > swap > device on the same IDE channel as another data storage device. Unlike > SCSI, IDE channels don't "overlay multiple i/o commands" or have > intelligent multiple outstanding I/O commands where the device > re-targets the host when > the I/O is completed. Furthermore, the "slave" device truly is a > second-class citizen, vis-a-vis contentious I/O access attempts. > Thanks for the detailed reply. I am not thinking this is a low level problem. The setup, as is, has been working under Windows just fine, ie the disk Mbytes/sec (up to ~17MB/s) and cpu utilization (a few %) for streaming is the same for the two drives, so the Master-Slave distinction does not seem to be important. With Cakewalk I can stream 8 stereo tracks off of the Slave drive without breaking a sweat. hdparm gives pretty much identical results for the two drives performance wise, going by your examples. They've been setup properly at boot time. Cable is right, bios settings, etc etc... It looks more like some kind of unecessary double buffering going on, as if data from hdb is being written to swap (or some buffer) first, then read from there by the application (XMMS or others). This is with Mandrake 10, but I wouldn't think it would be specific to that distro. I'm stumped. Seems like others would have run into this at some point. AC
