Placement Case for 30 Bins

This is a demonstration of the initial object placement using my Allocation Optimization Model. It is a parallel, heuristic, fail-safe model that runs as a pipeline.

The designed data allocation and replication model can place subsets of a data population into distributed systems such as computer clusters and grids. I perform a stochastic computer network simulation with up to 10,000 jobs and 100,000 data points.

I demonstrate data allocation using different replication factors and various numbers of subsets in order to optimize the computer network. The goal in this example is to minimize network delays and waiting times.

Note on replication factor: Copying 40% of a subset leads to a replication factor of 1.4 for 30 subsets (or targets). These bins are abstract locations that represent sets with assigned capacities.