[EMAIL PROTECTED] wrote:
I think we need a clearer definition of the problem and whether we, given the constraints of the current VM system, can actually achieve knowing this metric; and if not, what alternative metric we do want. And whether we want the new VM system to do such accounting properly.
In the current VM system this is an unsolvable problem from a performance standpoint. There are tools that exist today that are used for capacity planning that use the "sum of all pmaps" approach and they work as well as the interface which isn't very. But it's as good as you're going to get on the current VM design which has no mechanism for tracking who is using which pages (the current VM was designed before shared memory became the rule rather than the exception, amongst many other things).
In the new VM system this problem will be not only tractable but is on our list of observability things to plan for.
- Eric _______________________________________________ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
