A strategy for scheduling tightly coupled parallel applications on clusters

Abstract

SUMMARYAlthough various strategies have been developed for scheduling parallel applications with independent tasks, very little work exists for scheduling tightly coupled parallel applications on cluster environments. In this paper, we compare four different strategies based on performance models of tightly coupled parallel applications for scheduling the applications on clusters. In addition to algorithms based on existing popular optimization techniques, we also propose a new algorithm called Box Elimination that searches the space of performance model parameters to determine the best schedule of machines. By means of real and simulation experiments, we evaluated the algorithms on single cluster and multi-cluster setups. We show that our Box Elimination algorithm generates up to 80% more efficient schedules than other algorithms. We also show that the execution times of the schedules produced by our algorithm are more robust against the performance modeling errors.

Process Rescheduling in High Performance Computing Environments

“…In addition, the same problem is shown in the following two works. Sanjay and Vadhiyar (Sanjay & Vadhiyar, 2009) present a scheduling algorithm called Box Elimination. It considers a 3-D box of CPU, bandwidth and processors tuples for selecting the resources with minimum available CPU and bandwidth.…”

“…Applications containing routines for solving linear systems and fast Fourier transform (FFT) are typical examples of tightly-coupled parallel applications that may take profit from the power of cluster-of-clusters (Sanjay & Vadhiyar, 2009). Given that these clusters can be heterogeneous and the links among them are normally not fast, scheduling and load balancing are two key techniques that must be organized to reach high performance in this architecture.…”

“…Certainly, this idea is time consuming and new applications and resources require a new series of tests. After developing the first version of MigBSP, we have observed the promotion of intelligent scheduling systems which adjust their parameters on the fly and hide intrinsic complexity and optimization decisions from users (Ding et al, 2009;Nascimento et al, 2007;Sanjay & Vadhiyar, 2009). In this context, we developed a new heuristic named AutoMig that selects one or more candidates for migration automatically.…”

Strategies for Rescheduling Tightly-Coupled Parallel Applications in Multi-Cluster Grids

“…Another previous work [11], discussed scheduling algorithms that use the predictions by the performance models of the earlier effort to choose a set of resources in a non-dedicated cluster for execution of a single-phase parallel application with uniform computation and communication characteristics throughout its execution. The focus of the current work is to propose rescheduling algorithms for adaptive execution of multi-phase parallel applications on single and multiple clusters.…”

Design and implementation of an adaptive job allocation strategy for heterogeneous multi‐cluster computing systems

“…Faster compute nodes in the cluster must wait for slower compute nodes and some cluster resources are wasted. One of the most important questions, therefore, is how jobs can be more efficiently processed by multiple clusters with varying capabilities [10][11][12][13][14][15][16].…”