A hybrid mapping and scheduling algorithm for distributed workflow applications in a heterogeneous computing environment

Mengxia Zhu, Fei Cao, Jia Mi

Research output: Chapter in Book/Report/Conference proceedingChapterpeer-review

3 Scopus citations


Computing intensive scientific workflows structured as a directed acyclic graph (DAG) are widely applied to various distributed science and engineering applications to enable efficient knowledge discovery by automated data processing. Effective mapping and scheduling the workflow modules to the underlying distributed computing environment with heterogeneous resources for optimal network performance has remained as a challenge and attracted research efforts with many simulations and real experiments carried out in the grid and cloud infrastructures. Due to the computing intractability of this type of optimization problem, heuristic algorithms are commonly proposed to achieve the minimum end-to-end delay (EED) or other objectives such as maximum reliability and stability. In this paper, a Hybrid mapping algorithm combining Recursive Critical Path search and layer-based Priority techniques (HRCPP) is designed and developed to achieve the minimum EED. Four representative mapping and scheduling algorithms for minimum EED are compared with HRCPP. Our simulation results illustrate that HRCPP consistently achieves the smallest EED with a low algorithm running time observed from many different scales of simulated test cases.

Original languageEnglish
Title of host publicationIntelligent Distributed Computing V
Subtitle of host publicationProceedings of the 5th International Symposium on Intelligent Distributed Computing - IDC 2011, Delft, The Netherlands - October 2011
EditorsFrances M.T. Brazier, Kees Nieuwenhuis, Gregor Pavlin, Martijn Warnier, Costin Badica
Number of pages11
StatePublished - 22 Nov 2011

Publication series

NameStudies in Computational Intelligence
ISSN (Print)1860-949X


Dive into the research topics of 'A hybrid mapping and scheduling algorithm for distributed workflow applications in a heterogeneous computing environment'. Together they form a unique fingerprint.

Cite this