> Something that caught their attention though was that > the number of I/Os on the mainframe were about 30 times (not percent) > higher than on Intel. They said they'd used an Oracle tool to determine > this. Not knowing anything about Oracle's tools in this arena, I have no > idea if they're any good or not. I told them they needed to get a real > performance monitor to verify that, but with a 5.6-to-1 performance > _improvement_, it's not yet critical.
Oracle tends to aggressively cache data in RAM when it can, doing I/O only when it can't avoid it -- which is why most distributed Oracle servers tend to have truly obscene quantities of RAM. If someone did the virtual machine sizing to minimize the size of the virtual machine (as any good virtual machine should do), then it's logical to expect the virtual machine to do more I/O to reflect that it isn't caching as much, and the System z I/O subsystem is just absorbing the difference, as designed. > My question is, does anyone know if Oracle uses different internal > algorithms on the various architectures to maximize performance? That is, > since I/O is a strength of the mainframe, they might do more I/Os versus > something else on Intel Linux. Anyone have any insight on this? Don't know for certain here, but several Oracle people have said in other forums that they try to keep the code as similar as possible for multiple platforms. The only platform that's radically different is z/OS; the rest are mostly the same. ---------------------------------------------------------------------- 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
