Two-Stage Memory Allocation using AHP & Knapsack at PT Berca Hardayaperkasa

We propose to manage a (MicroStrategy) Business Intelligence Server in terms of RAM allocation for its Intelligent Cubes as a two-stage resource allocation problem in which the first stage is formulated as an multi-criteria problem that can be solved using Analytic Hierarchy Process (AHP) and the second stage is multiple (several) 0-1 classic Knapsack problems with the constraints that are
obtained using the result from the first stage. This Approach happens to have an advantage in terms of computational complexity as well, it reduces from O(nM) to O(max{nj}max{Mj}) when calculated in parallel. We illustrate our proposal with a numerical example based on our experience.