Title :
PTG-based parallel scheduling algorithm research on grid DAG task
Author :
Hong, LIANG ; Jie, Chen
Author_Institution :
Coll. of Comput. & Commun. Eng., China Univ. of Pet., Dongying, China
fDate :
June 29 2010-July 1 2010
Abstract :
This paper introduced a novel scheduling algorithm for grid DAG task, and built a tree level scheduling model. Through turning the grid DAG task into PTG (Parameterized Task Graph), adopting online and batch combination method that uses online processing to each layer and batch processing to tasks in layers. Through the experiment of simulation platform, the algorithm can effectively improve the parallel efficiency, and reduce the scheduling time.
Keywords :
batch processing (computers); directed graphs; grid computing; parallel algorithms; scheduling; task analysis; trees (mathematics); PTG based parallel scheduling algorithm; batch combination method; batch processing; directed cycling graph; grid DAG task; online processing; parameterized task graph; scheduling time; tree level scheduling model; Algorithm design and analysis; Computational modeling; Processor scheduling; DAG; Parameterized Task Graph(PTG); grid; parallel;
Conference_Titel :
Communication Systems, Networks and Applications (ICCSNA), 2010 Second International Conference on
Conference_Location :
Hong Kong
Print_ISBN :
978-1-4244-7475-2
DOI :
10.1109/ICCSNA.2010.5588716
