Framework For Dynamic Load Balancing Research Articles

ON TUNING THE DYNAMIC LOAD BALANCING FEM FRAMEWORK

Developments in computer hardware are currently bringing new opportunities for numerical modelling. The current trend in technology is parallel processing making use of multiple processing units simultaneously to solve a given problem. This paper deals with exploring the parallel dynamic load balancing framework implemented in the finite element software. This framework is based on a domain decomposition paradigm for distributed memory model. The paper describes the improved technique to determine the actual processor weights related to performance of individual processing units. The load recovery consisting in mesh (re)partitioning is based on actual processor weights. The (re)partitioning process has to be performed during the simulation and whenever the load imbalance is significant. The performance of the proposed technique is tested on the benchmark problem and discussed.

The Mathematical Theory of Dynamic Load Balancing in Cellular Networks

While many interesting dynamic load balancing schemes have been proposed for efficient use of limited bandwidth and to increase the capacity of congested or hot spots (or cells) in wireless networks, to date, a comprehensive mathematical framework which encompasses all of these schemes does not exist. In this paper, we provide a unified mathematical framework for dynamic load balancing, which leads to closed-form performance expressions for evaluating the performance of some of the most important dynamic load balancing strategies proposed in the literature. To the best of our knowledge, this is the first generic theoretical framework that can be used to evaluate the performance of many different dynamic load balancing schemes with simple closed-form results. The accuracy of the results predicted by these analytical expressions derived from the theoretical framework is checked by comparing these results with simulation results provided in the literature for well-known schemes.

Broader dynamic load balancing for hybrid/multi-level parallel programming models

Recently, Hybrid/Multi-Level Parallel Programming Models have begun gaining lots of momentum basically because they have proven to provide better scalability, speedup and utilisation than any single parallel programming model alone. In such models, load balancing should not only mean balancing the computational loads (as it has always been perceived), but also balancing I/O imbalance as well as synchronisation imbalance. In this paper, we propose a broader generic application/language/model independent multi-agent framework for dynamic load balancing. It takes most of the load-balancing burden away from programmers. It is not a library but a runtime support system that is not hardwired to the parallel applications. The framework is intended to handle varying levels of load changes in computations, I/O and/or synchronisation throughout the application run and it is an open architecture that currently supports four multi-level parallel programming models. It has a clean interface to the application, runs in parallel and provides additional functionality such as determination of when to load balance, and provide interface to end users. The proposed framework has been deployed in four hybrid/multi-level parallel programming models and its capabilities of issuing corrective actions against emerging imbalances were tested in the context of an adaptive mesh refinement application. Experimental results show that the framework is effective in monitoring, tuning and rebalancing emerging computational, I/O and synchronisation sources of load imbalance.

A Locality-conscious load-balancer based on negotiations in dynamic unstructured mesh computations

Recently hybrid/multi-level parallel programming models are gaining lots of momentum basically because they have proven to provide better scalability, speedup and utilization than any single parallel programming model alone. In such models, load balancing should not only mean balancing the computational loads (as it has always been perceived), but should also mean balancing I/O imbalance as well as synchronization imbalance. In this paper, we propose a broader generic application/language/model independent multi-agent framework for dynamic load balancing. It takes most of the load-balancing burden away from programmers. It is not a library but a runtime support system that is not hardwired to the parallel applications. The framework is intended to handle varying levels of load changes in computations, I/O and/or synchronization throughout the application run and it is an open-architecture that currently supports four multi-level parallel programming models. It has a clean interface to the application, runs in parallel and provides additional functionality such as determination of when to balance load and provide interface to end users. The proposed open-architecture multi-agent load-balancing capability currently makes use of a leading geometric partitioner engine (Chaco) at runtime. A mesh solver may initially create hundreds of lightweight threads, each handling a small submesh by calling Chaco partitioning engine in a pre-processing stage. This partitioner engine might be called again by these light-weight threads if a divide-and-conquer process is deemed necessary when the sub-domain (submesh) served by this thread grows out beyond certain threshold limits and thus creates an imbalance. In the proposed framework, the multi-agent is a set of SMP-based load balancers (agents) that do not have to share any data structure with the parallel application threads. They just monitor and collect system and application data frequently from the outside of the multi-threaded parallel application solver and send adjustments and negotiation plans to the SMP-load balancers and the application threads whenever a need for load balancing arises. The proposed framework has been deployed in four hybrid/multi-level parallel programming models and its capabilities of issuing corrective actions against emerging imbalances were tested in the context of an adaptive mesh refinement application. Experimental results show that the framework is effective in monitoring, tuning and rebalancing emerging computational, I/O and synchronization sources of load imbalance.

AN ADVERSARIAL MODEL FOR DISTRIBUTED DYNAMIC LOAD BALANCING

We study the problem of balancing the load on processors in an arbitrary network. If jobs arrive or depart during the process of load balancing, we have the dynamic load balancing problem; otherwise, we have the static load balancing problem. While static load balancing on a number of different networks has been well-studied, analytical work on dynamic load balancing is limited. The difficulty lies in modeling the arrivals and departures of jobs in a clean manner. In this paper, we introduce a new framework for dynamic load balancing in which the job traffic is modeled by an adversary. Our main result is that a simple local load balancing algorithm maintains the load of the network within a stable level against a powerful adversary. Our result holds for different models of traffic patterns and processor communication.

Parallel tetrahedral mesh adaptation with dynamic load balancing

The ability to dynamically adapt an unstructured-grid is a powerful tool for efficiently solving computational problems with evolving physical features. In this paper, we report on our experience parallelizing an edge-based adaptation scheme, called 3D_TAG, using message passing. Results show excellent speedup when a realistic helicopter rotor mesh is randomly refined. However, performance deteriorates when the mesh is refined using a solution-based error indicator since mesh adaptation for practical problems occurs in a localized region, creating a severe load imbalance. To address this problem, we have developed PLUM, a global dynamic load balancing framework for adaptive numerical computations. Even though PLUM primarily balances processor workloads for the solution phase, it reduces the load imbalance problem within mesh adaptation by repartitioning the mesh after targeting edges for refinement but before the actual subdivision. This dramatically improves the performance of parallel 3D_TAG since refinement occurs in a more load balanced fashion. We also present optimal and heuristic algorithms that, when applied to the default mapping of a parallel repartitioner, significantly reduce the data redistribution overhead. Finally, portability is examined by comparing performance on three state-of-the-art parallel machines.

A framework for dynamic load balancing: A case study on explosive containment simulation

Methods that can circumvent the implementation of parallelization steps are usually warmly welcomed by parallel simulation architects. In this paper we apply the parallel cellular automata modeling environment (P-CAM) to a parallel finite element simulation. P-CAM adopts a particle-based framework in which we can instantiate a variety of simulation problems. However, it can also be used to monitor and steer a parallel application outside this framework. A meta-model of the finite element grid is used within the framework to dynamically repartition the grid during the simulation. Based on this meta model a dynamic load balancing process, implemented in P-CAM, decides on a possible local repartitioning of the mesh. The system is successfully used to simulate highly dynamic transient behavior of an explosive detonating under a water surface.

HARP: A Dynamic Spectral Partitioner

Partitioning unstructured graphs is central to the parallel solution of computational science and engineering problems. Spectral partitioners, such as recursive spectral bisection (RSB), have proven effective in generating high-quality partitions of realistically sized meshes. The major problem which hindered their widespread use was their long execution times. This paper presents a new inertial spectral partitioner called HARP. The main objective of the proposed approach is to quickly partition the meshes at runtime for the dynamic load balancing framework JOVE which dynamically balances the computational loads of distributed-memory machines with a global view. The underlying principle of HARP is to find the eigenvectors of the unpartitioned vertices and then project them onto the eigenvectors of the original mesh. Results for various meshes ranging in size from 1000 to 100,000 vertices indicate that HARP can indeed partition meshes rapidly at runtime. Experimental results show that our largest mesh can be partitioned sequentially in only a few seconds on an SP-2, which is several times faster than other spectral partitioners, while maintaining the solution quality of the proven RSB method. These results indicate that graph partitioning can now be truly embedded in dynamically changing real-world applications.