>From: zapcat <zapcat at speakeasy.net> >On Jun 17, 2008, at 10:22 PM, David Ledger wrote: >> It doesn't gauge performance, it's a cause of poor performance. All it >> tells you about load is that you're asking your system to do more than >> you have the RAM for. > >well, if it "tells you" something, isn't that what a gauge >does....informs you of something? Well, in English English, 'gauge' means 'measure' or quantity information. Lots of pageouts shows you would probably benefit from more memory. That's not a measure, or anything like it. Just an indication. Of course if it's different in your English ... Following on from other's comments, pageins are not a problem. They happen every time an application starts. If there's free memory, the page data just gets read in. One operation. Pageouts are done to innocent bystander processes when some other process needs to pagein and there's no free memory. It's an extra largeish procedure that has to be gone through just because memory is full. Pageouts are never done otherwise. (If memory-mapped file access is used, and I don't know if it is in OSX, and if those file read and write operations are counted into the same stats because they are identical at a low level, then pagein and pageout stats will be skewed). David - currently in Las Vegas at HP Technology Forum, having just been to a presentation where exactly this was discussed as it applies to HP-UX. -- David Ledger - Freelance Unix Sysadmin in the UK. HP-UX specialist of hpUG technical user group (www.hpug.org.uk) david.ledger at ivdcs.co.uk www.ivdcs.co.uk