Recipes for Baking Black Forest Databases: Building and Querying Black Hole Merger Trees from Cosmological Simulations

Carnegie Mellon University Parallel Data Lab Technical Report CMU-PDL-11-104. April 2011.

Julio López, Colin Degraf, Tiziana DiMatteo, Bin Fu

Carnegie Mellon University
Pittsburgh, PA 15213


Large-scale N-body simulations play an important role in advancing our understanding of the formation and evolution of large structures in the universe. These computations require a large number of particles, in the order of 10-100 of billions, to realistically model phenomena such as the formation of galaxies. Among these particles, black holes play a dominant role on the formation of these structure. Computational cosmologists are interested in the analysis of black hole properties throughout the simulation with high temporal resolution. The properties of the black holes need to be assembled in merger tree histories to model the process where two or more black holes merge to form a larger one. In the past this analysis has been carried out with custom approaches that no longer scales to the size of black hole datasets produced by current cosmological simulations. We present a set of algorithms and a strategy to represent and store a forest of merger trees for black holes in relational databases (RDBMS). We implemented this approach and present results with datasets containing 0.5 billion time series records belonging to over 2 million black holes. We demonstrate that this is a feasible approach to support interactive analysis and enables flexible exploration of black hole forest datasets.

FULL TR: pdf




© 2017. Last updated 15 March, 2012