Xing Gao, Zhang Xu, Haining Wang, Li Li, Xiaorui Wang
A trend in modern data centers is to raise the temperature and maintain all servers in a relatively hot environment. While this can save on cooling costs given benign workloads running in servers, the hot environment increases the risk of cooling failure. In this work, we introduce the security concept of thermal attack on a data center that exploits thermal-intensive workloads to severely worsen the thermal conditions in the data center. To unveil the vulnerability of a data center to thermal attacks, we conduct thermal measurements and propose effective thermal attack vectors. To evaluate the impacts of thermal attacks inside a data center, we simulate datacenter-level thermal attacks using a real-world data center trace. Our evaluation demonstrates that thermal attacks can cause local hotspots, and even worse lead to cooling failures.
{"title":"Why \"Some\" Like It Hot Too: Thermal Attack on Data Centers","authors":"Xing Gao, Zhang Xu, Haining Wang, Li Li, Xiaorui Wang","doi":"10.1145/3078505.3078545","DOIUrl":"https://doi.org/10.1145/3078505.3078545","url":null,"abstract":"A trend in modern data centers is to raise the temperature and maintain all servers in a relatively hot environment. While this can save on cooling costs given benign workloads running in servers, the hot environment increases the risk of cooling failure. In this work, we introduce the security concept of thermal attack on a data center that exploits thermal-intensive workloads to severely worsen the thermal conditions in the data center. To unveil the vulnerability of a data center to thermal attacks, we conduct thermal measurements and propose effective thermal attack vectors. To evaluate the impacts of thermal attacks inside a data center, we simulate datacenter-level thermal attacks using a real-world data center trace. Our evaluation demonstrates that thermal attacks can cause local hotspots, and even worse lead to cooling failures.","PeriodicalId":133673,"journal":{"name":"Proceedings of the 2017 ACM SIGMETRICS / International Conference on Measurement and Modeling of Computer Systems","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121948480","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Session details: Session 7: Resource Allocation & Economics","authors":"Lei Ying","doi":"10.1145/3248543","DOIUrl":"https://doi.org/10.1145/3248543","url":null,"abstract":"","PeriodicalId":133673,"journal":{"name":"Proceedings of the 2017 ACM SIGMETRICS / International Conference on Measurement and Modeling of Computer Systems","volume":"517 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123102868","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
N. Wu, Yingjie Bi, Nithin Michael, A. Tang, J. Doyle, N. Matni
We propose high-frequency traffic control (HFTraC), a rate control scheme that coordinates the transmission rates and buffer utilizations in routers network-wide at fast timescale. HFTraC can effectively deal with traffic demand fluctuation by utilizing available buffer space in routers network-wide, and therefore lead to significant performance improvement in terms of tradeoff between bandwidth utilization and queueing delay. We further note that the performance limit of HFTraC is determined by the network architecture used to implement it. We provide trace-driven evaluation of the performance of HFTraC implemented in the proposed architectures that vary from fully centralized to completely decentralized.
{"title":"HFTraC: High-Frequency Traffic Control","authors":"N. Wu, Yingjie Bi, Nithin Michael, A. Tang, J. Doyle, N. Matni","doi":"10.1145/3078505.3078557","DOIUrl":"https://doi.org/10.1145/3078505.3078557","url":null,"abstract":"We propose high-frequency traffic control (HFTraC), a rate control scheme that coordinates the transmission rates and buffer utilizations in routers network-wide at fast timescale. HFTraC can effectively deal with traffic demand fluctuation by utilizing available buffer space in routers network-wide, and therefore lead to significant performance improvement in terms of tradeoff between bandwidth utilization and queueing delay. We further note that the performance limit of HFTraC is determined by the network architecture used to implement it. We provide trace-driven evaluation of the performance of HFTraC implemented in the proposed architectures that vary from fully centralized to completely decentralized.","PeriodicalId":133673,"journal":{"name":"Proceedings of the 2017 ACM SIGMETRICS / International Conference on Measurement and Modeling of Computer Systems","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122601648","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Session details: Session 6: New Design for Large Network Applications","authors":"Anshul Gandhi","doi":"10.1145/3248541","DOIUrl":"https://doi.org/10.1145/3248541","url":null,"abstract":"","PeriodicalId":133673,"journal":{"name":"Proceedings of the 2017 ACM SIGMETRICS / International Conference on Measurement and Modeling of Computer Systems","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125206528","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Mean-field analysis is an analytical method for understanding large-scale stochastic systems such as large-scale data centers and communication networks. The idea is to approximate the stationary distribution of a large-scale stochastic system using the equilibrium point (called the mean-field limit) of a dynamical system (called the mean-field model). This approximation is often justified by proving the weak convergence of stationary distributions to its mean-field limit. Most existing mean-field models concerned the light-traffic regime where the load of the system, denote by ρ, is strictly less than one and is independent of the size of the system. This is because a traditional mean-field model represents the limit of the corresponding stochastic system. Therefore, the load of the mean-field model is ρ=limN-> ∞ ρ(N), where ρ(N) is the load of the stochastic system of size N. Now if ρ(N)-> 1 as N -> ∞ (i.e., in the heavy-traffic regime), then ρ=1. For most systems, the mean-field limits when ρ=1 are trivial and meaningless. To overcome this difficulty of traditional mean-field models, this paper takes a different point of view on mean-field models. Instead of regarding a mean-field model as the limiting system of large-scale stochastic system, it views the equilibrium point of the mean-field model, called a mean-field solution, simply as an approximation of the stationary distribution of the finite-size system. Therefore both mean-field models and solutions can be functions of N. The proposed method focuses on quantifying the approximation error. If the approximation error is small (as we will show in two applications), then we can conclude that the mean-field solution is a good approximation of the stationary distribution.
{"title":"Stein's Method for Mean-Field Approximations in Light and Heavy Traffic Regimes","authors":"Lei Ying","doi":"10.1145/3078505.3078592","DOIUrl":"https://doi.org/10.1145/3078505.3078592","url":null,"abstract":"Mean-field analysis is an analytical method for understanding large-scale stochastic systems such as large-scale data centers and communication networks. The idea is to approximate the stationary distribution of a large-scale stochastic system using the equilibrium point (called the mean-field limit) of a dynamical system (called the mean-field model). This approximation is often justified by proving the weak convergence of stationary distributions to its mean-field limit. Most existing mean-field models concerned the light-traffic regime where the load of the system, denote by ρ, is strictly less than one and is independent of the size of the system. This is because a traditional mean-field model represents the limit of the corresponding stochastic system. Therefore, the load of the mean-field model is ρ=limN-> ∞ ρ(N), where ρ(N) is the load of the stochastic system of size N. Now if ρ(N)-> 1 as N -> ∞ (i.e., in the heavy-traffic regime), then ρ=1. For most systems, the mean-field limits when ρ=1 are trivial and meaningless. To overcome this difficulty of traditional mean-field models, this paper takes a different point of view on mean-field models. Instead of regarding a mean-field model as the limiting system of large-scale stochastic system, it views the equilibrium point of the mean-field model, called a mean-field solution, simply as an approximation of the stationary distribution of the finite-size system. Therefore both mean-field models and solutions can be functions of N. The proposed method focuses on quantifying the approximation error. If the approximation error is small (as we will show in two applications), then we can conclude that the mean-field solution is a good approximation of the stationary distribution.","PeriodicalId":133673,"journal":{"name":"Proceedings of the 2017 ACM SIGMETRICS / International Conference on Measurement and Modeling of Computer Systems","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130635630","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Multi-Path TCP (MPTCP) is a new standardized transport protocol that enables devices to utilize multiple network interfaces. The default MPTCP path scheduler prioritizes paths with the smallest round trip time (RTT). In this work, we examine whether the default MPTCP path scheduler can provide applications the ideal aggregate bandwidth, i.e., the sum of available bandwidths of all paths. Our experimental results show that heterogeneous paths cause under-utilization of the fast path, resulting in undesirable application behaviors such as lower video streaming quality than can be obtained using the available aggregate bandwidth. To solve this problem, we propose and implement a new MPTCP path scheduler, ECF (Earliest Completion First), that utilizes all relevant information about a path, not just RTT. Our results show that ECF consistently utilizes all available paths more efficiently than other approaches under path heterogeneity, particularly for streaming video.
{"title":"ECF: An MPTCP Path Scheduler to Manage Heterogeneous Paths","authors":"Yeon-sup Lim, E. Nahum, D. Towsley, R. Gibbens","doi":"10.1145/3078505.3078552","DOIUrl":"https://doi.org/10.1145/3078505.3078552","url":null,"abstract":"Multi-Path TCP (MPTCP) is a new standardized transport protocol that enables devices to utilize multiple network interfaces. The default MPTCP path scheduler prioritizes paths with the smallest round trip time (RTT). In this work, we examine whether the default MPTCP path scheduler can provide applications the ideal aggregate bandwidth, i.e., the sum of available bandwidths of all paths. Our experimental results show that heterogeneous paths cause under-utilization of the fast path, resulting in undesirable application behaviors such as lower video streaming quality than can be obtained using the available aggregate bandwidth. To solve this problem, we propose and implement a new MPTCP path scheduler, ECF (Earliest Completion First), that utilizes all relevant information about a path, not just RTT. Our results show that ECF consistently utilizes all available paths more efficiently than other approaches under path heterogeneity, particularly for streaming video.","PeriodicalId":133673,"journal":{"name":"Proceedings of the 2017 ACM SIGMETRICS / International Conference on Measurement and Modeling of Computer Systems","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130761089","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Session details: Session 4: Performance Analysis of Very Large Systems","authors":"Mor Harchol-Balter","doi":"10.1145/3248539","DOIUrl":"https://doi.org/10.1145/3248539","url":null,"abstract":"","PeriodicalId":133673,"journal":{"name":"Proceedings of the 2017 ACM SIGMETRICS / International Conference on Measurement and Modeling of Computer Systems","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120950851","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Highly dynamic vehicular networks use long-range radio technologies such as DSRC, WiMAX, and Cellular networks to maintain connectivity. Multipath TCP offers the possibility to combine these radio technologies to improve network performance, allow robust handoffs, and maintain vehicle connectivity at all times. The proliferation of mobile devices with dual interfaces and the manufacturers' interest to make their vehicles smarter and more competitive create the ideal scenario for MPTCP on VANETs. In this paper, we study the performance of MPTCP on two VANET scenarios: Vehicle-to-Infrastructure (V2I), and Vehicle-to-Vehicle, (V2V) under distinct velocities.
{"title":"Multipath TCP on a VANET: A Performance Study","authors":"Jorge Mena, Peter Bankole, M. Gerla","doi":"10.1145/3078505.3078555","DOIUrl":"https://doi.org/10.1145/3078505.3078555","url":null,"abstract":"Highly dynamic vehicular networks use long-range radio technologies such as DSRC, WiMAX, and Cellular networks to maintain connectivity. Multipath TCP offers the possibility to combine these radio technologies to improve network performance, allow robust handoffs, and maintain vehicle connectivity at all times. The proliferation of mobile devices with dual interfaces and the manufacturers' interest to make their vehicles smarter and more competitive create the ideal scenario for MPTCP on VANETs. In this paper, we study the performance of MPTCP on two VANET scenarios: Vehicle-to-Infrastructure (V2I), and Vehicle-to-Vehicle, (V2V) under distinct velocities.","PeriodicalId":133673,"journal":{"name":"Proceedings of the 2017 ACM SIGMETRICS / International Conference on Measurement and Modeling of Computer Systems","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125505987","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Principal Component Analysis (PCA) and Canonical Component Analysis (CCA) are two of the few examples of non-convex optimization problems that can be solved efficiently with sharp guarantees. This is achieved by the classical and well-established understanding of matrix factorizations. Recently, several new theoretical and algorithmic challenges have arisen in statistical learning over matrix factorizations, motivated by various real-world applications. Despite the inherent non-convex nature of these problem, efficient algorithms are being discovered with provable guarantees, extending the frontier of our understanding of non-convex optimization problems. I will present several recent results in this area in applications to matrix completion and sensing, crowdsourcing, ranking, and tensor factorization.
{"title":"Matrix Factorization at the Frontier of Non-convex Optimizations: Abstract for SIGMETRICS 2017 Rising Star Award Talk","authors":"Sewoong Oh","doi":"10.1145/3143314.3080573","DOIUrl":"https://doi.org/10.1145/3143314.3080573","url":null,"abstract":"Principal Component Analysis (PCA) and Canonical Component Analysis (CCA) are two of the few examples of non-convex optimization problems that can be solved efficiently with sharp guarantees. This is achieved by the classical and well-established understanding of matrix factorizations. Recently, several new theoretical and algorithmic challenges have arisen in statistical learning over matrix factorizations, motivated by various real-world applications. Despite the inherent non-convex nature of these problem, efficient algorithms are being discovered with provable guarantees, extending the frontier of our understanding of non-convex optimization problems. I will present several recent results in this area in applications to matrix completion and sensing, crowdsourcing, ranking, and tensor factorization.","PeriodicalId":133673,"journal":{"name":"Proceedings of the 2017 ACM SIGMETRICS / International Conference on Measurement and Modeling of Computer Systems","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125696818","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Siddharth Singh, Vedant Nanda, Rijurekha Sen, Sohaib Ahmad, Satadal Sengupta, A. Phokeer, Z. Farooq, T. Khan, P. Kumaraguru, I. Qazi, D. Choffnes, K. Gummadi
ACM Reference format: Siddharth Singh, Vedant Nanda, Rijurekha Sen, Sohaib Ahmad, Satadal Sengupta, Amreesh Phokeer, Zaid Ahmed Farooq, Taslim Arefin Khan, Ponnurangam Kumaraguru, Ihsan Ayyub Qazi, David Choffnes, and Krishna P. Gummadi. 2017. An Empirical Analysis of Facebook’s Free Basics Program. In Proceedings of ACM Sigmetrics conference, Urbana-Champaign,IL, USA, June 05-09, 2017 (SIGMETRICS ’17), 2 pages. https://doi.org/http://dx.doi.org/10.1145/3078505.3078554
ACM参考格式:Siddharth Singh, Vedant Nanda, Rijurekha Sen, Sohaib Ahmad, Satadal Sengupta, Amreesh Phokeer, Zaid Ahmed Farooq, Taslim Arefin Khan, Ponnurangam Kumaraguru, Ihsan Ayyub Qazi, David Choffnes和Krishna P. Gummadi, 2017。Facebook免费基础项目的实证分析ACM Sigmetrics会议论文集，Urbana-Champaign,IL, USA, June 05-09, 2017 (Sigmetrics ' 17)， 2页。https://doi.org/http://dx.doi.org/10.1145/3078505.3078554
{"title":"An Empirical Analysis of Facebook's Free Basics","authors":"Siddharth Singh, Vedant Nanda, Rijurekha Sen, Sohaib Ahmad, Satadal Sengupta, A. Phokeer, Z. Farooq, T. Khan, P. Kumaraguru, I. Qazi, D. Choffnes, K. Gummadi","doi":"10.1145/3078505.3078554","DOIUrl":"https://doi.org/10.1145/3078505.3078554","url":null,"abstract":"ACM Reference format: Siddharth Singh, Vedant Nanda, Rijurekha Sen, Sohaib Ahmad, Satadal Sengupta, Amreesh Phokeer, Zaid Ahmed Farooq, Taslim Arefin Khan, Ponnurangam Kumaraguru, Ihsan Ayyub Qazi, David Choffnes, and Krishna P. Gummadi. 2017. An Empirical Analysis of Facebook’s Free Basics Program. In Proceedings of ACM Sigmetrics conference, Urbana-Champaign,IL, USA, June 05-09, 2017 (SIGMETRICS ’17), 2 pages. https://doi.org/http://dx.doi.org/10.1145/3078505.3078554","PeriodicalId":133673,"journal":{"name":"Proceedings of the 2017 ACM SIGMETRICS / International Conference on Measurement and Modeling of Computer Systems","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129025906","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：