Lee Ann Riesen

Hypergraph-based Dynamic Load Balancing for Adaptive Scientific Computations

Adaptive scientific computations require that periodic repartitioning (load balancing) occur dynamically to maintain load balance. Hypergraph partitioning is a successful model for minimizing communication volume in scientific computations, and partitioning software for the static case is widely available. In this paper, we present a new hypergraph model for the dynamic case, where we minimize the sum of communication in the application plus the migration cost to move data, thereby reducing total execution time. The new model can be solved using hypergraph partitioning with faced vertices. We describe an implementation of a parallel multilevel repartitioning algorithm within the Zoltan load-balancing toolkit, which to our knowledge is the first code for dynamic load balancing based on hypergraph partitioning. Finally, we present experimental results that demonstrate the effectiveness of our approach on a Linux cluster with up to 64 processors. Our new algorithm compares favorably to the widely used ParMETIS partitioning software in terms of quality, and would have reduced total execution time in most of our test cases.

A repartitioning hypergraph model for dynamic load balancing

In parallel adaptive applications, the computational structure of the applications changes over time, leading to load imbalances even though the initial load distributions were balanced. To restore balance and to keep communication volume low in further iterations of the applications, dynamic load balancing (repartitioning) of the changed computational structure is required. Repartitioning differs from static load balancing (partitioning) due to the additional requirement of minimizing migration cost to move data from an existing partition to a new partition. In this paper, we present a novel repartitioning hypergraph model for dynamic load balancing that accounts for both communication volume in the application and migration cost to move data, in order to minimize the overall cost. The use of a hypergraph-based model allows us to accurately model communication costs rather than approximate them with graph-based models. We show that the new model can be realized using hypergraph partitioning with fixed vertices and describe our parallel multilevel implementation within the Zoltan load balancing toolkit. To the best of our knowledge, this is the first implementation for dynamic load balancing based on hypergraph partitioning. To demonstrate the effectiveness of our approach, we conducted experiments on a Linux cluster with 1024 processors. The results show that, in terms of reducing total cost, our new model compares favorably to the graph-based dynamic load balancing approaches, and multilevel approaches improve the repartitioning quality significantly.

US Environmental Protection Agency Uses Operations Research to Reduce Contamination Risks in Drinking Water

The US Environmental Protection Agency (EPA) is the lead federal agency for the security of drinking water in the United States. The agency is responsible for providing information and technical assistance to the more than 50,000 water utilities across the country. The distributed physical layout of drinking-water utilities makes them inherently vulnerable to contamination incidents caused by terrorists. To counter this threat, the EPA is using operations research to design, test, and deploy contamination warning systems (CWSs) that rapidly detect the presence of contaminants in drinking water. We developed a software tool to optimize the design process, published a decision-making process to assist utilities in applying the tool, pilot-tested the tool on nine large water utilities, and provided training and technical assistance to a larger group of utilities. We formed a collaborative team of industry, academia, and government to critique our approach and share CWS deployment experiences. Our work has demonstrated that a CWS is a cost-effective, timely, and capable method of detecting a broad range of contaminants. Widespread application of these new systems will significantly reduce the risks associated with catastrophic contamination incidents: the median estimated fatalities reduction for the nine utilities already studied is 48 percent; the corresponding economic-impact reduction is over

The TEVA-SPOT Toolkit for Drinking Water Contaminant Warning System Design

19 billion. Because of this operations research program, online monitoring programs, such as a CWS, are now the accepted technology for reducing contamination risks in drinking water.

Limited-Memory Techniques for Sensor Placement in Water Distribution Networks

We present the TEVA-SPOT Toolkit, a sensor placement optimization tool developed within the USEPA TEVA program. The TEVA-SPOT Toolkit provides a sensor placement framework that facilitates research in sensor placement optimization and enables the practical application of sensor placement solvers to real-world CWS design applications. This paper provides an overview of its key features, and then illustrates how this tool can be flexibly applied to solve a variety of different types of sensor placement problems.

Low-memory Lagrangian Relaxation Methods for Sensor Placement in Municipal Water Networks.

The practical utility of optimization technologies is often impacted by factors that reflect how these tools are used in practice, including whether various real-world constraints can be adequately modeled, the sophistication of the analysts applying the optimizer, and related environmental factors (e.g. whether a company is willing to trust predictions from computational models). Other features are less appreciated, but of equal importance in terms of dictating the successful use of optimization. These include the scale of problem instances, which in practice drives the development of approximate solution techniques, and constraints imposed by the target computing platforms. End-users often lack state-of-the-art computers, and thus runtime and memory limitations are often a significant, limiting factor in algorithm design. When coupled with large problem scale, the result is a significant technological challenge. We describe our experience developing and deploying both exact and heuristic algorithms for placing sensors in water distribution networks to mitigate against damage due intentional or accidental introduction of contaminants. The target computing platforms for this application have motivated limited-memory techniques that can optimize large-scale sensor placement problems.

Getting Started with Zoltan: a Short Tutorial ?

Placing sensors in municipal water networks to protect against a set of contamination events is a classic p-median problem for most objectives when we assume that sensors are perfect. Many researchers have proposed exact and approximate solution methods for this p-median formulation. For full-scale networks with large contamination event suites, one must generally rely on heuristic methods to generate solutions. These heuristics provide feasible solutions, but give no quality guarantee relative to the optimal placement. In this paper we apply a Lagrangian relaxation method in order to compute lower bounds on the expected impact of suites of contamination events. In all of our experiments with single objectives, these lower bounds establish that the GRASP local search method generates solutions that are provably optimal to to within a fraction of a percentage point. Our Lagrangian heuristic also provides good solutions itself and requires only a fraction of the memory of GRASP. We conclude by describing two variations of the Lagrangian heuristic: an aggregated version that trades off solution quality for further memory savings, and a multiobjective version which balances objectives with additional goals.