ACM Transactions on Modeling and Performance Evaluation of Computing Systems (TOMPECS)最新文献

{"title":"GPSonflow","authors":"Elizabeth Varki","doi":"10.1145/3197656","DOIUrl":"https://doi.org/10.1145/3197656","url":null,"abstract":"This article evaluates the maximum data flow from a sender to a receiver via the internet when all transmissions are scheduled for early morning hours. The significance of early morning hours is that internet congestion is low while users sleep. When the sender and receiver lie in proximal time zones, a direct transmission from sender to receiver can be scheduled for early morning hours. When the sender and receiver are separated by several time zones such that their sleep times are non-overlapping, data can still be transmitted during early morning hours with an indirect store-and-forward transfer. The data are transmitted from the sender to intermediate end networks or data centers that serve as storage hops en route to receiver. The storage hops are placed in zones that are time proximal to the sender or the receiver so that all transmissions to and from storage hops occur during low-congestion early morning hours. This article finds the optimal locations and bandwidth distributions of storage hops for maximum nice internet flow from a sender to a receiver.","PeriodicalId":105474,"journal":{"name":"ACM Transactions on Modeling and Performance Evaluation of Computing Systems (TOMPECS)","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115526693","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":"Efficiency and Optimality of Largest Deficit First Prioritization","authors":"Yuhuan Du, G. Veciana","doi":"10.1145/3200479","DOIUrl":"https://doi.org/10.1145/3200479","url":null,"abstract":"An increasing number of real-time applications with compute and/or communication deadlines are being supported on a shared infrastructure. Such applications can often tolerate occasional deadline violations without substantially impacting their Quality of Service (QoS). This article explores the efficient allocation of shared resources to satisfy such QoS requirements. We study a simple framework which decouples this problem into two subproblems: (1) dynamic prioritization of users based on arbitrary functions of their deficits (difference of achieved vs. required QoS), and (2) priority-based resource allocation on the underlying compute fabric. In this article, we shall assume the solution to the latter is fixed, e.g., as realized in the task prioritization capabilities of current hardware/software, and focus on compatible solutions to the former. We first characterize the set of feasible QoS requirements and show the optimality of max weight-like prioritization. We then consider simple weighted Largest Deficit First (w-LDF) prioritization policies, where users with higher weighted QoS deficits are given higher priority. The article gives an inner bound for the feasible set under w-LDF policies, and, under an additional monotonicity assumption, characterizes its geometry leading to a sufficient condition for optimality. Additional insights on the efficiency ratio of w-LDF policies, the optimality of hierarchical-LDF, and characterization of clustering of failures are also discussed.","PeriodicalId":105474,"journal":{"name":"ACM Transactions on Modeling and Performance Evaluation of Computing Systems (TOMPECS)","volume":"81 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134191916","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":"Searching for a Single Community in a Graph","authors":"Avik Ray, S. Sanghavi, S. Shakkottai","doi":"10.1145/3200863","DOIUrl":"https://doi.org/10.1145/3200863","url":null,"abstract":"In standard graph clustering/community detection, one is interested in partitioning the graph into more densely connected subsets of nodes. In contrast, the search problem of this article aims to only find the nodes in a single such community, the target, out of the many communities that may exist. To do so, we are given suitable side information about the target; for example, a very small number of nodes from the target are labeled as such. We consider a general yet simple notion of side information: all nodes are assumed to have random weights, with nodes in the target having higher weights on average. Given these weights and the graph, we develop a variant of the method of moments that identifies nodes in the target more reliably, and with lower computation, than generic community detection methods that do not use side information and partition the entire graph. Our empirical results show significant gains in runtime, and also gains in accuracy over other graph clustering algorithms.","PeriodicalId":105474,"journal":{"name":"ACM Transactions on Modeling and Performance Evaluation of Computing Systems (TOMPECS)","volume":"83 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124724755","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":"An Experimental Performance Evaluation of Autoscalers for Complex Workflows","authors":"A. Ilyushkin, A. Ali-Eldin, N. Herbst, A. Bauer, A. Papadopoulos, D. Epema, A. Iosup","doi":"10.1145/3164537","DOIUrl":"https://doi.org/10.1145/3164537","url":null,"abstract":"Elasticity is one of the main features of cloud computing allowing customers to scale their resources based on the workload. Many autoscalers have been proposed in the past decade to decide on behalf of cloud customers when and how to provision resources to a cloud application based on the workload utilizing cloud elasticity features. However, in prior work, when a new policy is proposed, it is seldom compared to the state-of-the-art, and is often compared only to static provisioning using a predefined quality of service target. This reduces the ability of cloud customers and of cloud operators to choose and deploy an autoscaling policy, as there is seldom enough analysis on the performance of the autoscalers in different operating conditions and with different applications. In our work, we conduct an experimental performance evaluation of autoscaling policies, using as application model workflows, a popular formalism for automating resource management for applications with well-defined yet complex structures. We present a detailed comparative study of general state-of-the-art autoscaling policies, along with two new workflow-specific policies. To understand the performance differences between the seven policies, we conduct various experiments and compare their performance in both pairwise and group comparisons. We report both individual and aggregated metrics. As many workflows have deadline requirements on the tasks, we study the effect of autoscaling on workflow deadlines. Additionally, we look into the effect of autoscaling on the accounted and hourly based charged costs, and we evaluate performance variability caused by the autoscaler selection for each group of workflow sizes. Our results highlight the trade-offs between the suggested policies, how they can impact meeting the deadlines, and how they perform in different operating conditions, thus enabling a better understanding of the current state-of-the-art.","PeriodicalId":105474,"journal":{"name":"ACM Transactions on Modeling and Performance Evaluation of Computing Systems (TOMPECS)","volume":"70 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127706793","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":"An Empirical Analysis of Amazon EC2 Spot Instance Features Affecting Cost-Effective Resource Procurement","authors":"Cheng Wang, Qianlin Liang, B. Urgaonkar","doi":"10.1145/3164538","DOIUrl":"https://doi.org/10.1145/3164538","url":null,"abstract":"Many cost-conscious public cloud workloads (“tenants”) are turning to Amazon EC2’s spot instances because, on average, these instances offer significantly lower prices (up to 10 times lower) than on-demand and reserved instances of comparable advertised resource capacities. To use spot instances effectively, a tenant must carefully weigh the lower costs of these instances against their poorer availability. Toward this, we empirically study four features of EC2 spot instance operation that a cost-conscious tenant may find useful to model. Using extensive evaluation based on historical spot instance data, we show shortcomings in the state-of-the-art modeling of these features that we overcome. As an extension to our prior work, we conduct data analysis on a rich dataset of the latest spot price traces collected from a variety of EC2 spot markets. Our analysis reveals many novel properties of spot instance operation, some of which offer predictive value whereas others do not. Using these insights, we design predictors for our features that offer a balance between computational efficiency (allowing for online resource procurement) and cost efficacy. We explore “case studies” wherein we implement prototypes of dynamic spot instance procurement advised by our predictors for two types of workloads. Compared to the state of the art, our approach achieves (i) comparable cost but much better performance (fewer bid failures) for a latency-sensitive in-memory Memcached cache and (ii) an additional 18% cost savings with comparable (if not better than) performance for a delay-tolerant batch workload.","PeriodicalId":105474,"journal":{"name":"ACM Transactions on Modeling and Performance Evaluation of Computing Systems (TOMPECS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122903760","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":"Special Issue","authors":"A. Koziolek, E. Smirni","doi":"10.1145/3186329","DOIUrl":"https://doi.org/10.1145/3186329","url":null,"abstract":"We are delighted to present in this special issue of TOMPECS three of the highest-rated papers of the Research Track of the 8th ACM/SPEC International Conference on Performance Engineering (ICPE) that took place in L’Aquila, Italy, April 22–26, 2017. The goal of ICPE is to integrate theory and practice in the field of performance engineering by providing a forum for sharing ideas and experiences between industry and academia. The conference grew out of the ACM Workshop on Software Performance (WOSP since 1998) and the SPEC International Performance Engineering Workshop (SIPEW since 2008). In the ICPE 2017 edition, the research track attracted 65 submissions with 22 accepted for presentation as full papers. The program covered traditional ICPE topics such as design for performance and problem diagnosis, online performance management, analytic models, empirical studies, model building, and benchmarking, as well as application of performance engineering theory and techniques to several practical fields, including distributed systems, cloud computing, storage, energy, big data, virtualized systems, and hardware. In this special issue, we present the extended versions of three of the highest-rated papers in the conference. Three members of the ICPE 2017 program committee served as reviewers for each extended version, and each paper went through the regular TOMPECS review/revision cycle. The first article by Cheng Wang (VMware Inc.), Qianlin Liang (Ele.me Inc.), and Bhuvan Urgaonkar (Penn State University), titled “An Empirical Analysis of Amazon EC2 Spot Instance Features Affecting Cost-Effective Resource Procurement,” focuses on ways to optimize the usage of Amazon’s popular EC2 spot markets. The article defines four features that can result in a better tenant bid: instance lifetime, average spot price during lifetime, simultaneous revocations, and startup delay. For these features, they illustrate how current prediction practices are inadequate and propose new and more accurate prediction alternatives. The proposed predictors are extensively evaluated using real-world workloads and show that their approach results in better performance (measured in diminished bid failures and lower tail latency) with the same cost as prior work. Results are corroborated with two sets of trace data: a 3-month period in 2015 and another 3-month period in 2017. The work is a case study in traditional ICPE topics: problem diagnosis and performance prediction in cloud computing.","PeriodicalId":105474,"journal":{"name":"ACM Transactions on Modeling and Performance Evaluation of Computing Systems (TOMPECS)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134132627","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":"RAPL in Action","authors":"Kashif Nizam Khan, Mikael Hirki, T. Niemi, J. Nurminen, Zhonghong Ou","doi":"10.1145/3177754","DOIUrl":"https://doi.org/10.1145/3177754","url":null,"abstract":"To improve energy efficiency and comply with the power budgets, it is important to be able to measure the power consumption of cloud computing servers. Intel’s Running Average Power Limit (RAPL) interface is a powerful tool for this purpose. RAPL provides power limiting features and accurate energy readings for CPUs and DRAM, which are easily accessible through different interfaces on large distributed computing systems. Since its introduction, RAPL has been used extensively in power measurement and modeling. However, the advantages and disadvantages of RAPL have not been well investigated yet. To fill this gap, we conduct a series of experiments to disclose the underlying strengths and weaknesses of the RAPL interface by using both customized microbenchmarks and three well-known application level benchmarks: Stream, Stress-ng, and ParFullCMS. Moreover, to make the analysis as realistic as possible, we leverage two production-level power measurement datasets from the Taito, a supercomputing cluster of the Finnish Center of Scientific Computing and also replicate our experiments on Amazon EC2. Our results illustrate different aspects of RAPL and document the findings through comprehensive analysis. Our observations reveal that RAPL readings are highly correlated with plug power, promisingly accurate enough, and have negligible performance overhead. Experimental results suggest RAPL can be a very useful tool to measure and monitor the energy consumption of servers without deploying any complex power meters. We also show that there are still some open issues, such as driver support, non-atomicity of register updates, and unpredictable timings that might weaken the usability of RAPL in certain scenarios. For such scenarios, we pinpoint solutions and workarounds.","PeriodicalId":105474,"journal":{"name":"ACM Transactions on Modeling and Performance Evaluation of Computing Systems (TOMPECS)","volume":"109 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115799088","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":"Disk Prefetching Mechanisms for Increasing HTTP Streaming Video Server Throughput","authors":"Benjamin Cassell, Tyler Szepesi, Jim Summers, Tim Brecht, D. Eager, B. Wong","doi":"10.1145/3164536","DOIUrl":"https://doi.org/10.1145/3164536","url":null,"abstract":"Most video streaming traffic is delivered over HTTP using standard web servers. While traditional web server workloads consist of requests that are primarily for small files that can be serviced from the file system cache, HTTP video streaming workloads often service a long tail of large infrequently requested videos. As a result, optimizing disk accesses is critical to obtaining good server throughput. In this article we explore serialized, aggressive disk prefetching, a technique that can be used to improve the throughput of HTTP streaming video web servers. We identify how serialization and aggressive prefetching affect performance, and, based on our findings, we construct and evaluate Libception, an application-level shim library that implements both techniques. By dynamically linking against Libception at runtime, applications are able to transparently obtain benefits from serialization and aggressive prefetching without needing to change their source code. In contrast to other approaches that modify applications, make kernel changes, or attempt to optimize kernel tuning, Libception provides a portable and relatively simple system in which techniques for optimizing I/O in HTTP video streaming servers can be implemented and evaluated. We empirically evaluate the efficacy of serialization and aggressive prefetching both with and without Libception, using three web servers (Apache, nginx, and the userver) running on two operating systems (FreeBSD and Linux). We find that, by using Libception, we can improve streaming throughput for all three web servers by at least a factor of 2 on FreeBSD and a factor of 2.5 on Linux. Additionally, we find that with significant tuning of Linux kernel parameters, we can achieve similar performance to Libception by globally modifying Linux’s disk prefetch behaviour. Finally, we demonstrate Libception’s ability to reduce the completion time of a microbenchmark involving two applications competing for disk resources.","PeriodicalId":105474,"journal":{"name":"ACM Transactions on Modeling and Performance Evaluation of Computing Systems (TOMPECS)","volume":"601 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116326414","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":"An Online Emergency Demand Response Mechanism for Cloud Computing","authors":"Ruiting Zhou, Zongpeng Li, Chuan Wu","doi":"10.1145/3177755","DOIUrl":"https://doi.org/10.1145/3177755","url":null,"abstract":"This article studies emergency demand response (EDR) mechanisms from a data center perspective, where a cloud participates in a mandatory EDR program while receiving computing job bids from cloud users in an online fashion. We target a realistic EDR mechanism where (i) the cloud provider dynamically packs different types of resources on servers into requested VMs and computes job schedules to meet users’ requirements, (ii) the power consumption of servers in the cloud is limited by the grid through the EDR program, and (iii) the operation cost of the cloud is considered in the calculation of social welfare, measured by an electricity cost that consists of both volume charge and peak charge. We propose an online auction for dynamic cloud resource provisioning that is under the control of the EDR program, runs in polynomial time, achieves truthfulness, and close-to-optimal social welfare for the cloud ecosystem. In the design of the online auction, we first propose a new framework, compact exponential LPs, to handle job scheduling constraints in the time domain. We then develop a posted pricing auction framework toward the truthful online auction design, which leverages the classic primal-dual technique for approximation algorithm design. We evaluate our online auctions through both theoretical analysis and empirical studies driven by real-world traces.","PeriodicalId":105474,"journal":{"name":"ACM Transactions on Modeling and Performance Evaluation of Computing Systems (TOMPECS)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115662708","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":"Bargaining Game-Based Scheduling for Performance Guarantees in Cloud Computing","authors":"Chubo Liu, Kenli Li, Zhuo Tang, Kuan-Ching Li","doi":"10.1145/3141233","DOIUrl":"https://doi.org/10.1145/3141233","url":null,"abstract":"In this article, we focus on request scheduling with performance guarantees of all users in cloud computing. Each cloud user submits requests with average response time requirement, and the cloud provider tries to find a scheduling scheme, i.e., allocating user requests to limited servers, such that the average response times of all cloud users can be guaranteed. We formulate the considered scenario into a cooperative game among multiple users and try to find a Nash bargaining solution (NBS), which can simultaneously satisfy all users’ performance demands. We first prove the existence of NBS and then analyze its computation. Specifically, for the situation when all allocating substreams are strictly positive, we propose a computational algorithm (CA), which can find the NBS very efficiently. For the more general case, we propose an iterative algorithm (IA), which is based on duality theory. The convergence of our proposed IA algorithm is also analyzed. Finally, we conduct some numerical calculations. The experimental results show that our IA algorithm can find an appropriate scheduling strategy and converges to a stable state very quickly.","PeriodicalId":105474,"journal":{"name":"ACM Transactions on Modeling and Performance Evaluation of Computing Systems (TOMPECS)","volume":"115 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132642807","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}