Standard

An Experimental Performance Evaluation of Autoscaling Policies for Complex Workflows. / Ilyushkin, Alexey; Ali-Eldin, Ahmed; Herbst, Nikolas; Herbst, Nikolas; Papadopoulos, Alessandro; Ghit, Bogdan; Epema, Dick; Iosup, Alexandru.

8th ACM/SPEC Int'l Conference in Performance Engineering (ICPE). ACM DL, 2017. p. 75.

Research output: Scientific - peer-reviewConference contribution

Harvard

Ilyushkin, A, Ali-Eldin, A, Herbst, N, Herbst, N, Papadopoulos, A, Ghit, B, Epema, D & Iosup, A 2017, An Experimental Performance Evaluation of Autoscaling Policies for Complex Workflows. in 8th ACM/SPEC Int'l Conference in Performance Engineering (ICPE). ACM DL, pp. 75. DOI: 10.1145/3030207.3030214

APA

Ilyushkin, A., Ali-Eldin, A., Herbst, N., Herbst, N., Papadopoulos, A., Ghit, B., ... Iosup, A. (2017). An Experimental Performance Evaluation of Autoscaling Policies for Complex Workflows. In 8th ACM/SPEC Int'l Conference in Performance Engineering (ICPE). (pp. 75). ACM DL. DOI: 10.1145/3030207.3030214

Vancouver

Ilyushkin A, Ali-Eldin A, Herbst N, Herbst N, Papadopoulos A, Ghit B et al. An Experimental Performance Evaluation of Autoscaling Policies for Complex Workflows. In 8th ACM/SPEC Int'l Conference in Performance Engineering (ICPE). ACM DL. 2017. p. 75. Available from, DOI: 10.1145/3030207.3030214

Author

Ilyushkin, Alexey; Ali-Eldin, Ahmed; Herbst, Nikolas; Herbst, Nikolas; Papadopoulos, Alessandro; Ghit, Bogdan; Epema, Dick; Iosup, Alexandru / An Experimental Performance Evaluation of Autoscaling Policies for Complex Workflows.

8th ACM/SPEC Int'l Conference in Performance Engineering (ICPE). ACM DL, 2017. p. 75.

Research output: Scientific - peer-reviewConference contribution

BibTeX

@inbook{72778db789aa4dc98105217a2d1d6f64,
title = "An Experimental Performance Evaluation of Autoscaling Policies for Complex Workflows",
author = "Alexey Ilyushkin and Ahmed Ali-Eldin and Nikolas Herbst and Nikolas Herbst and Alessandro Papadopoulos and Bogdan Ghit and Dick Epema and Alexandru Iosup",
year = "2017",
doi = "10.1145/3030207.3030214",
pages = "75",
booktitle = "8th ACM/SPEC Int'l Conference in Performance Engineering (ICPE)",
publisher = "ACM DL",

}

RIS

TY - CHAP

T1 - An Experimental Performance Evaluation of Autoscaling Policies for Complex Workflows

AU - Ilyushkin,Alexey

AU - Ali-Eldin,Ahmed

AU - Herbst,Nikolas

AU - Herbst,Nikolas

AU - Papadopoulos,Alessandro

AU - Ghit,Bogdan

AU - Epema,Dick

AU - Iosup,Alexandru

PY - 2017

Y1 - 2017

N2 - Simplifying the task of resource management and scheduling forcustomers, while still delivering complex Quality-of-Service (QoS),is key to cloud computing. Many autoscaling policies have beenproposed in the past decade to decide on behalf of cloud customerswhen and how to provision resources to a cloud application utilizingcloud elasticity features. However, in prior work, when a new policyis proposed, it is seldom compared to the state-of-the-art, and isoften compared only to static provisioning using a predefined QoStarget. This reduces the ability of cloud customers and of cloudoperators to choose and deploy an autoscaling policy. In our work,we conduct an experimental performance evaluation of autoscalingpolicies, using as application model workflows, a commonly usedformalism for automating resource management for applicationswith well-defined yet complex structure. We present a detailedcomparative study of general state-of-the-art autoscaling policies,along with two new workflow-specific policies. To understand theperformance differences between the 7 policies, we conduct variousforms of pairwise and group comparisons. We report both individualand aggregated metrics. Our results highlight the trade-offs betweenthe suggested policies, and thus enable a better understanding of the current state-of-the-art.

AB - Simplifying the task of resource management and scheduling forcustomers, while still delivering complex Quality-of-Service (QoS),is key to cloud computing. Many autoscaling policies have beenproposed in the past decade to decide on behalf of cloud customerswhen and how to provision resources to a cloud application utilizingcloud elasticity features. However, in prior work, when a new policyis proposed, it is seldom compared to the state-of-the-art, and isoften compared only to static provisioning using a predefined QoStarget. This reduces the ability of cloud customers and of cloudoperators to choose and deploy an autoscaling policy. In our work,we conduct an experimental performance evaluation of autoscalingpolicies, using as application model workflows, a commonly usedformalism for automating resource management for applicationswith well-defined yet complex structure. We present a detailedcomparative study of general state-of-the-art autoscaling policies,along with two new workflow-specific policies. To understand theperformance differences between the 7 policies, we conduct variousforms of pairwise and group comparisons. We report both individualand aggregated metrics. Our results highlight the trade-offs betweenthe suggested policies, and thus enable a better understanding of the current state-of-the-art.

U2 - 10.1145/3030207.3030214

DO - 10.1145/3030207.3030214

M3 - Conference contribution

SP - 75

BT - 8th ACM/SPEC Int'l Conference in Performance Engineering (ICPE)

PB - ACM DL

ER -

ID: 29617131