Exploring the Performance Fluctuations of HPC Workloads on Clouds

Author

El-Khamra, Yaakoub ; Kim, Hyunjoo ; Jha, Shantenu ; Parashar, Manish

Author_Institution

Texas Adv. Comput. Center, Univ. of Texas at Austin, Austin, TX, USA

fYear

2010

fDate

Nov. 30 2010-Dec. 3 2010

Firstpage

383

Lastpage

387

Abstract

Clouds enable novel execution modes often supported by advanced capabilities such as autonomic schedulers. These capabilities are predicated upon an accurate estimation and calculation of runtimes on a given infrastructure. Using a well understood high-performance computing workload, we find strong fluctuations from the mean performance on EC2 and Eucalyptus-based cloud systems. Our analysis eliminates variations in IO and computational times as possible causes, we find that variations in communication times account for the bulk of the experiment-to-experiment fluctuations of the performance.

Keywords

cloud computing; fault tolerant computing; performance evaluation; EC2; Eucalyptus-based cloud systems; HPC workloads; autonomic schedulers; clouds infrastructure; experiment-to-experiment fluctuations; high-performance computing workload; performance fluctuations; Analytical models; Benchmark testing; Cloud computing; Conferences; Hardware; Radio access networks; Runtime; Cloud Computing; Concurrent Programming; Distributed Programming; Parallel Programming;

fLanguage

English

Publisher

ieee

Conference_Titel

Cloud Computing Technology and Science (CloudCom), 2010 IEEE Second International Conference on

Conference_Location

Indianapolis, IN

Print_ISBN

978-1-4244-9405-7

Electronic_ISBN

978-0-7695-4302-4

Type

conf

DOI

10.1109/CloudCom.2010.84

Filename

5708474