PRM Question and understanding

[Mark Killen]

I am getting ready to implement prm on a dev server and have completed my reading and some testing on a r&d server but would just like to ensure that I am understand one thing correctly. Will prm turned on with a cpu limitation of 40% cpu to two different groups, allow each group to get more cpu if the other is not using it and throttle it back when the 2 apps are busy.
I understand the capping and was able to test that.
Many Thanks

[Todd McDaniel_1 · Accepted solution]

If I remember correctly, if you give 40% to each app, they will not give up any of that but the 20% left over can be used as variable by either.

I attached the PRM guide...

[Todd McDaniel_1]

There is a way to reserve the 40% but not restrict it for a certain app. So that if one is using less than 40 it will give it up to the pool of cpu free... but if it needs it back, it will take it back and wean the cpu back to its primary application.

I havent used it in a while to remember. I think that is correct.

[Ted Buis]

The default policy is to not invoke the resource rules, until there is a constraint. However, it is possible to set the policy so that the limitations are enforced even if there is no constraint.