Title :
ExoApp: Performance Evaluation of Data-Intensive Applications on ExoGENI
Author :
Ze Yu ; Xinxin Liu ; Min Li ; Kaikai Liu ; Xiaolin Li
Author_Institution :
Scalable Software Syst. Lab., Univ. of Florida, Gainesville, FL, USA
Abstract :
ExoGENI is a new GENI-federated Infrastructureas- a-Service (IaaS) framework. In this paper, we evaluate the performance of data-intensive applications on ExoGENI´s resources. To simplify experiments, we design an automatic provisioning system called ExoApp. This paper focuses on MapReduce-based applications. Users can easily deploy applications in ExoGENI using ExoApp, without having to manually configure cluster runtime environments. We then conduct a series of experiments using real-world data sets and standard benchmarks through ExoApp. Our result shows that ExoGENI demonstrates similar resource quality when hosting data-intensive applications and its Network-as-a-Service (NaaS) model maintains stable network performance. We finally identify the pros and cons of the ExoGENI´s NaaS model in supporting data-intensive applications.
Keywords :
Web services; cloud computing; parallel processing; software performance evaluation; ExoApp; ExoGENI; GENI-federated infrastructureas-a-service framework; IaaS framework; MapReduce-based applications; NaaS model; Web service; automatic provisioning system; cluster runtime environments; data-intensive applications; network-as-a-service; performance evaluation; resource quality; Bandwidth; Benchmark testing; Biological system modeling; Cloud computing; Computational modeling; Programming; Runtime; MapReduce; Experiment; Data-intensive; ExoGENI;
Conference_Titel :
Research and Educational Experiment Workshop (GREE), 2013 Second GENI
Conference_Location :
Salt Lake City, UT
DOI :
10.1109/GREE.2013.14