Research Article
Sensitivity of Application Performance to Resource Availability
@INPROCEEDINGS{10.4108/eai.25-10-2016.2266602, author={Ajitha Rajan and Boris Penev}, title={Sensitivity of Application Performance to Resource Availability}, proceedings={10th EAI International Conference on Performance Evaluation Methodologies and Tools}, publisher={ACM}, proceedings_a={VALUETOOLS}, year={2017}, month={5}, keywords={sensitivity index resource contention monte carlo analysis}, doi={10.4108/eai.25-10-2016.2266602} }- Ajitha Rajan
Boris Penev
Year: 2017
Sensitivity of Application Performance to Resource Availability
VALUETOOLS
ACM
DOI: 10.4108/eai.25-10-2016.2266602
Abstract
Existing literature has extensively explored and analysed the effects of shared resource contention to improve system throughput and resource utilisation. Nevertheless, a systematic study varying different resource availabilities and examining their combined effect on the performance of an individual application has not been conducted earlier. In this paper, we perform global sensitivity analysis using Monte Carlo simulations of three resource parameters -- CPU availability, RAM availability and number of cores, and examine their effect on execution time of industry standard benchmark programs. In addition to understanding application performance sensitivity, our study also lends itself
to identifying threshold levels of required resources, below which severe application performance degradation occurs.