umu.sePublications
Change search
Link to record
Permanent link

Direct link
BETA
Elmroth, Erik
Alternative names
Publications (10 of 179) Show all publications
Vu, X.-S., Addi, A.-M., Elmroth, E. & Lili, J. (2019). Graph-based Interactive Data Federation System for Heterogeneous Data Retrieval and Analytics. In: Proceedings of The 30th TheWebConf'19 (formerly WWW), USA: . Paper presented at The Web Conference, San Fransisco, USA, May 13-17, 2019 (pp. 3595-3599). New York, NY, USA: ACM Digital Library
Open this publication in new window or tab >>Graph-based Interactive Data Federation System for Heterogeneous Data Retrieval and Analytics
2019 (English)In: Proceedings of The 30th TheWebConf'19 (formerly WWW), USA, New York, NY, USA: ACM Digital Library, 2019, p. 3595-3599Conference paper, Published paper (Refereed)
Abstract [en]

Given the increasing number of heterogeneous data stored in relational databases, file systems or cloud environment, it needs to be easily accessed and semantically connected for further data analytic. The potential of data federation is largely untapped, this paper presents an interactive data federation system (https://vimeo.com/ 319473546) by applying large-scale techniques including heterogeneous data federation, natural language processing, association rules and semantic web to perform data retrieval and analytics on social network data. The system first creates a Virtual Database (VDB) to virtually integrate data from multiple data sources. Next, a RDF generator is built to unify data, together with SPARQL queries, to support semantic data search over the processed text data by natural language processing (NLP). Association rule analysis is used to discover the patterns and recognize the most important co-occurrences of variables from multiple data sources. The system demonstrates how it facilitates interactive data analytic towards different application scenarios (e.g., sentiment analysis, privacyconcern analysis, community detection).

Place, publisher, year, edition, pages
New York, NY, USA: ACM Digital Library, 2019
Keywords
heterogeneous data federation, RDF, interactive data analysis
National Category
Language Technology (Computational Linguistics)
Identifiers
urn:nbn:se:umu:diva-160892 (URN)10.1145/3308558.3314138 (DOI)978-1-4503-6674-8 (ISBN)
Conference
The Web Conference, San Fransisco, USA, May 13-17, 2019
Available from: 2019-06-25 Created: 2019-06-25 Last updated: 2019-08-22Bibliographically approved
Nguyen, C., Klein, C. & Elmroth, E. (2019). Multivariate LSTM-Based Location-Aware Workload Prediction for Edge Data Centers. In: Proceedings, 19th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing: . Paper presented at CCGrid 2019, 19th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (IEEE/ACM CCGrid 2019), 14-17 May, Larnaca, Cyprus (pp. 341-350). IEEE
Open this publication in new window or tab >>Multivariate LSTM-Based Location-Aware Workload Prediction for Edge Data Centers
2019 (English)In: Proceedings, 19th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, IEEE, 2019, p. 341-350Conference paper, Published paper (Refereed)
Abstract [en]

Mobile Edge Clouds (MECs) is a promising computing platform to overcome challenges for the success of bandwidth-hungry, latency-critical applications by distributing computing and storage capacity in the edge of the network as Edge Data Centers (EDCs) within the close vicinity of end-users. Due to the heterogeneous distributed resource capacity in EDCs, the application deployment flexibility coupled with the user mobility, MECs bring significant challenges to control resource allocation and provisioning. In order to develop a self-managed system for MECs which efficiently decides how much and when to activate scaling, where to place and migrate services, it is crucial to predict its workload characteristics, including variations over time and locality. To this end, we present a novel location-aware workload predictor for EDCs. Our approach leverages the correlation among workloads of EDCs in a close physical distance and applies multivariate Long Short-Term Memory network to achieve on-line workload predictions for each EDC. The experiments with two real mobility traces show that our proposed approach can achieve better prediction accuracy than a state-of-the art location-unaware method (up to 44%) and a location-aware method (up to 17%). Further, through an intensive performance measurement using various input shaking methods, we substantiate that the proposed approach achieves a reliable and consistent performance.

Place, publisher, year, edition, pages
IEEE, 2019
Keywords
Mobile Edge Cloud, Edge Data Center, ResourceManagement, Workload Prediction, Location-aware, MachineLearning
National Category
Computer Systems
Research subject
Computer Systems
Identifiers
urn:nbn:se:umu:diva-159540 (URN)10.1109/CCGRID.2019.00048 (DOI)000483058700039 ()2-s2.0-85069517887 (Scopus ID)978-1-7281-0912-1 (ISBN)978-1-7281-0913-8 (ISBN)
Conference
CCGrid 2019, 19th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (IEEE/ACM CCGrid 2019), 14-17 May, Larnaca, Cyprus
Available from: 2019-05-30 Created: 2019-05-30 Last updated: 2019-09-27Bibliographically approved
Nguyen, C. L., Mehta, A., Klein, C. & Elmroth, E. (2019). Why Cloud Applications Are not Ready for the Edge (yet). In: 4th ACM/IEEE Symposium on Edge Computing: . Paper presented at 4th ACM/IEEE Symposium on Edge Computing (SEC 2019). IEEE
Open this publication in new window or tab >>Why Cloud Applications Are not Ready for the Edge (yet)
2019 (English)In: 4th ACM/IEEE Symposium on Edge Computing, IEEE, 2019Conference paper, Published paper (Other academic)
Abstract [en]

Mobile Edge Clouds (MECs) are distributed platforms in which distant data-centers are complemented with computing and storage capacity located at the edge of the network. With such high resource distribution, MECs potentially fulfill the need of low latency and high bandwidth to offer an improved user experience.

As modern cloud applications are increasingly architected as collections of small, independently deployable services, it enables them to be flexibly deployed in various configurations combining resources from both centralized datacenters and edge location. Therefore, one might expect them to be well-placed to benefit from the advantage of MECs in order to reduce the service response time.In this paper, we quantify the benefits of deploying such cloud micro-service applications on MECs. Using two popular benchmarks, we show that, against conventional wisdom, end-to-end latency does not improve significantly even when most application services are deployed in the edge location. We developed a profiler to better understand this phenomenon, allowing us to develop recommendations for adapting applications to MECs. Further, by quantifying the gains of those recommendations, we show that the performance of an application can be made to reach the ideal scenario, in which the latency between an edge datacenter and a remote datacenter has no impact on the application performance.

This work thus presents ways of adapting cloud-native applications to take advantage of MECs and provides guidance for developing MEC-native applications. We believe that both these elements are necessary to drive MEC adoption.

Place, publisher, year, edition, pages
IEEE, 2019
Keywords
Mobile Edge Clouds, Edge Latency, Mobile Application Development, Micro-service, Profiling
National Category
Computer Sciences
Identifiers
urn:nbn:se:umu:diva-162930 (URN)
Conference
4th ACM/IEEE Symposium on Edge Computing (SEC 2019)
Funder
Wallenberg AI, Autonomous Systems and Software Program (WASP)
Available from: 2019-09-02 Created: 2019-09-02 Last updated: 2019-09-04
Ibidunmoye, O., Ali-Reza, R. & Elmroth, E. (2018). Adaptive Anomaly Detection in Performance Metric Streams. IEEE Transactions on Network and Service Management, 15(1), 217-231
Open this publication in new window or tab >>Adaptive Anomaly Detection in Performance Metric Streams
2018 (English)In: IEEE Transactions on Network and Service Management, ISSN 1932-4537, E-ISSN 1932-4537, Vol. 15, no 1, p. 217-231Article in journal (Refereed) Published
Abstract [en]

Continuous detection of performance anomalies such as service degradations has become critical in cloud and Internet services due to impact on quality of service and end-user experience. However, the volume and fast changing behaviour of metric streams have rendered it a challenging task. Many diagnosis frameworks often rely on thresholding with stationarity or normality assumption, or on complex models requiring extensive offline training. Such techniques are known to be prone to spurious false-alarms in online settings as metric streams undergo rapid contextual changes from known baselines. Hence, we propose two unsupervised incremental techniques following a two-step strategy. First, we estimate an underlying temporal property of the stream via adaptive learning and, then we apply statistically robust control charts to recognize deviations. We evaluated our techniques by replaying over 40 time-series streams from the Yahoo! Webscope S5 datasets as well as 4 other traces of real web service QoS and ISP traffic measurements. Our methods achieve high detection accuracy and few false-alarms, and better performance in general compared to an open-source package for time-series anomaly detection.

Place, publisher, year, edition, pages
IEEE, 2018
Keywords
Performance Monitoring and Measurement, Computer Network Management, Quality of Service, Time Series Analysis, Anomaly Detection, Unsupervised Learning
National Category
Computer Systems
Research subject
Computer Science; Computing Science; Computer Systems
Identifiers
urn:nbn:se:umu:diva-142030 (URN)10.1109/TNSM.2017.2750906 (DOI)000427420100016 ()
Projects
Cloud Control
Funder
Swedish Research Council, C0590801
Available from: 2017-11-17 Created: 2017-11-17 Last updated: 2018-08-07Bibliographically approved
Krzywda, J., Ali-Eldin, A., Wadbro, E., Östberg, P.-O. & Elmroth, E. (2018). ALPACA: Application Performance Aware Server Power Capping. In: ICAC 2018: 2018 IEEE International Conference on Autonomic Computing (ICAC), Trento, Italy, September 3-7, 2018. Paper presented at 15th IEEE International Conference on Autonomic Computing (ICAC 2018) (pp. 41-50). IEEE Computer Society
Open this publication in new window or tab >>ALPACA: Application Performance Aware Server Power Capping
Show others...
2018 (English)In: ICAC 2018: 2018 IEEE International Conference on Autonomic Computing (ICAC), Trento, Italy, September 3-7, 2018, IEEE Computer Society, 2018, p. 41-50Conference paper, Published paper (Refereed)
Abstract [en]

Server power capping limits the power consumption of a server to not exceed a specific power budget. This allows data center operators to reduce the peak power consumption at the cost of performance degradation of hosted applications. Previous work on server power capping rarely considers Quality-of-Service (QoS) requirements of consolidated services when enforcing the power budget. In this paper, we introduce ALPACA, a framework to reduce QoS violations and overall application performance degradation for consolidated services. ALPACA reduces unnecessary high power consumption when there is no performance gain, and divides the power among the running services in a way that reduces the overall QoS degradation when the power is scarce. We evaluate ALPACA using four applications: MediaWiki, SysBench, Sock Shop, and CloudSuite’s Web Search benchmark. Our experiments show that ALPACA reduces the operational costs of QoS penalties and electricity by up to 40% compared to a non optimized system. 

Place, publisher, year, edition, pages
IEEE Computer Society, 2018
Series
IEEE Conference Publication, ISSN 2474-0756
Keywords
power capping, performance degradation, power-performance tradeoffs
National Category
Computer Systems
Research subject
business data processing
Identifiers
urn:nbn:se:umu:diva-132428 (URN)10.1109/ICAC.2018.00014 (DOI)978-1-5386-5139-1 (ISBN)
Conference
15th IEEE International Conference on Autonomic Computing (ICAC 2018)
Available from: 2017-03-13 Created: 2017-03-13 Last updated: 2019-08-07Bibliographically approved
Mehta, A. & Elmroth, E. (2018). Distributed Cost-Optimized Placement for Latency-Critical Applications in Heterogeneous Environments. In: Proceedings of the IEEE 15th International Conference on Autonomic Computing (ICAC): . Paper presented at 2018 IEEE International Conference on Autonomic Computing, Trento, Italy, September 3-7, 2018 (pp. 121-130). IEEE Computer Society
Open this publication in new window or tab >>Distributed Cost-Optimized Placement for Latency-Critical Applications in Heterogeneous Environments
2018 (English)In: Proceedings of the IEEE 15th International Conference on Autonomic Computing (ICAC), IEEE Computer Society, 2018, p. 121-130Conference paper, Published paper (Refereed)
Abstract [en]

Mobile Edge Clouds (MECs) with 5G will create new opportunities to develop latency-critical applications in domains such as intelligent transportation systems, process automation, and smart grids. However, it is not clear how one can costefficiently deploy and manage a large number of such applications given the heterogeneity of devices, application performance requirements, and workloads. This work explores cost and performance dynamics for IoT applications, and proposes distributed algorithms for automatic deployment of IoT applications in heterogeneous environments. Placement algorithms were evaluated with respect to metrics including number of required runtimes, applications’ slowdown, and the number of iterations used to place an application. Iterative search-based distributed algorithms such as Size Interval Actor Assignment in Groups (SIAA G) outperformed random and bin packing algorithms, and are therefore recommended for this purpose. Size Interval Actor Assignment in Groups at Least Utilized Runtime (SIAA G LUR) algorithm is also recommended when minimizing the number of iterations is important. The tradeoff of using SIAA G algorithms is a few extra runtimes compared to bin packing algorithms.

Place, publisher, year, edition, pages
IEEE Computer Society, 2018
Series
Proceedings of the International Conference on Autonomic Computing, ISSN 2474-0764
Keywords
Mobile Edge Clouds, Fog Computing, IoTs, Distributed algorithms
National Category
Computer Systems
Identifiers
urn:nbn:se:umu:diva-151457 (URN)10.1109/ICAC.2018.00022 (DOI)978-1-5386-5139-1 (ISBN)
Conference
2018 IEEE International Conference on Autonomic Computing, Trento, Italy, September 3-7, 2018
Available from: 2018-09-04 Created: 2018-09-04 Last updated: 2019-06-26Bibliographically approved
Karakostas, V., Goumas, G., Bayuh Lakew, E., Elmroth, E., Gerangelos, S., Kolberg, S., . . . Koziris, N. (2018). Efficient Resource Management for Data Centers: The ACTiCLOUD Approach. In: Mudge T., Pnevmatikatos D.N. (Ed.), 2018 International conference on embedded computer systems: architectures, modeling, and simulation (SAMOS XVIII). Paper presented at SAMOS XVIII, July 15–19, 2018, Pythagorion, Samos Island, Greece (pp. 244-246). Association for Computing Machinery (ACM)
Open this publication in new window or tab >>Efficient Resource Management for Data Centers: The ACTiCLOUD Approach
Show others...
2018 (English)In: 2018 International conference on embedded computer systems: architectures, modeling, and simulation (SAMOS XVIII) / [ed] Mudge T., Pnevmatikatos D.N., Association for Computing Machinery (ACM), 2018, p. 244-246Conference paper, Published paper (Refereed)
Abstract [en]

Despite their proliferation as a dominant computing paradigm, cloud computing systems lack effective mechanisms to manage their vast resources efficiently. Resources are stranded and fragmented, limiting cloud applicability only to classes of applications that pose moderate resource demands. In addition, the need for reduced cost through consolidation introduces performance interference, as multiple VMs are co-located on the same nodes. To avoid such issues, current providers follow a rather conservative approach regarding resource management that leads to significant underutilization. ACTiCLOUD is a three-year Horizon 2020 project that aims at creating a novel cloud architecture that breaks existing scale-up and share-nothing barriers and enables the holistic management of physical resources, at both local and distributed cloud site levels. This extended abstract provides a brief overview of the resource management part of ACTiCLOUD, focusing on the design principles and the components.

Place, publisher, year, edition, pages
Association for Computing Machinery (ACM), 2018
Series
ACM International Conference Proceeding Series
Keywords
resource management, resource efficiency, cloud computing, data centers, in-memory databases, NUMA, heterogeneous, scale-up/out
National Category
Computer Systems
Identifiers
urn:nbn:se:umu:diva-154435 (URN)10.1145/3229631.3236095 (DOI)000475843000033 ()2-s2.0-85060997517 (Scopus ID)978-1-4503-6494-2 (ISBN)
Conference
SAMOS XVIII, July 15–19, 2018, Pythagorion, Samos Island, Greece
Available from: 2018-12-18 Created: 2018-12-18 Last updated: 2019-09-05Bibliographically approved
Bhuyan, M. H. & Elmroth, E. (2018). Multi-Scale Low-Rate DDoS Attack Detection Using the Generalized Total Variation Metric. In: 17th IEEE International Conference on Machine Learning and Applications: . Paper presented at 17th IEEE International Conference on Machine Learning and Applications, 2018, 17-20 December, Orlando, FL, USA (pp. 1040-1047). IEEE
Open this publication in new window or tab >>Multi-Scale Low-Rate DDoS Attack Detection Using the Generalized Total Variation Metric
2018 (English)In: 17th IEEE International Conference on Machine Learning and Applications, IEEE, 2018, p. 1040-1047Conference paper, Published paper (Refereed)
Abstract [en]

We propose a mechanism to detect multi-scale low-rate DDoS attacks which uses a generalized total variation metric. The proposed metric is highly sensitive towards detecting different variations in the network traffic and evoke more distance between legitimate and attack traffic as compared to the other detection mechanisms. Most low-rate attackers invade the security system by scale-in-and-out of periodic packet burst towards the bottleneck router which severely degrades the Quality of Service (QoS) of TCP applications. Our proposed mechanism can effectively identify attack traffic of this natures, despite its similarity to legitimate traffic, based on the spacing value of our metric. We evaluated our mechanism using datasets from CAIDA DDoS, MIT Lincoln Lab, and real-time testbed traffic. Our results demonstrate that our mechanism exhibits good accuracy and scalability in the detection of multi-scale low-rate DDoS attacks.

Place, publisher, year, edition, pages
IEEE, 2018
Keywords
Multi-scale, Distributed denial of service, Low-rate, Total variation metric
National Category
Computer Sciences
Research subject
Computer and Information Science
Identifiers
urn:nbn:se:umu:diva-155560 (URN)10.1109/ICMLA.2018.00170 (DOI)978-1-5386-6805-4 (ISBN)
Conference
17th IEEE International Conference on Machine Learning and Applications, 2018, 17-20 December, Orlando, FL, USA
Funder
The Kempe Foundations, SMK-1644
Available from: 2019-01-22 Created: 2019-01-22 Last updated: 2019-01-22Bibliographically approved
Krzywda, J., Ali-Eldin, A., Carlson, T. E., Östberg, P.-O. & Elmroth, E. (2018). Power-performance tradeoffs in data center servers: DVFS, CPUpinning, horizontal, and vertical scaling. Future generations computer systems, 81, 114-128
Open this publication in new window or tab >>Power-performance tradeoffs in data center servers: DVFS, CPUpinning, horizontal, and vertical scaling
Show others...
2018 (English)In: Future generations computer systems, ISSN 0167-739X, E-ISSN 1872-7115, Vol. 81, p. 114-128Article in journal (Refereed) Published
Abstract [en]

Dynamic Voltage and Frequency Scaling (DVFS), CPU pinning, horizontal, and vertical scaling, are four techniques that have been proposed as actuators to control the performance and energy consumption on data center servers. This work investigates the utility of these four actuators, and quantifies the power-performance tradeoffs associated with them. Using replicas of the German Wikipedia running on our local testbed, we perform a set of experiments to quantify the influence of DVFS, vertical and horizontal scaling, and CPU pinning on end-to-end response time (average and tail), throughput, and power consumption with different workloads. Results of the experiments show that DVFS rarely reduces the power consumption of underloaded servers by more than 5%, but it can be used to limit the maximal power consumption of a saturated server by up to 20% (at a cost of performance degradation). CPU pinning reduces the power consumption of underloaded server (by up to 7%) at the cost of performance degradation, which can be limited by choosing an appropriate CPU pinning scheme. Horizontal and vertical scaling improves both the average and tail response time, but the improvement is not proportional to the amount of resources added. The load balancing strategy has a big impact on the tail response time of horizontally scaled applications.

Keywords
Power-performance tradeoffs, Dynamic Voltage and Frequency Scaling (DVFS), CPU pinning, Horizontal scaling, Vertical scaling
National Category
Computer Systems
Research subject
Computer Science
Identifiers
urn:nbn:se:umu:diva-132427 (URN)10.1016/j.future.2017.10.044 (DOI)000423652200010 ()2-s2.0-85033772481 (Scopus ID)
Note

Originally published in thesis in manuscript form.

Available from: 2017-03-13 Created: 2017-03-13 Last updated: 2019-07-02Bibliographically approved
Gonzalo P., R., Elmroth, E., Östberg, P.-O. & Ramakrishnan, L. (2018). ScSF: a scheduling simulation framework. In: Proceedings of the 21th Workshop on Job Scheduling Strategies for Parallel Processing: . Paper presented at 21th Workshop on Job Scheduling Strategies for Parallel Processing (JSSP 2017), Orlando FL, USA, June 2nd, 2017 (pp. 152-173). Springer, 10773
Open this publication in new window or tab >>ScSF: a scheduling simulation framework
2018 (English)In: Proceedings of the 21th Workshop on Job Scheduling Strategies for Parallel Processing, Springer, 2018, Vol. 10773, p. 152-173Conference paper, Published paper (Refereed)
Abstract [en]

High-throughput and data-intensive applications are increasingly present, often composed as workflows, in the workloads of current HPC systems. At the same time, trends for future HPC systems point towards more heterogeneous systems with deeper I/O and memory hierarchies. However, current HPC schedulers are designed to support classical large tightly coupled parallel jobs over homogeneous systems. Therefore, There is an urgent need to investigate new scheduling algorithms that can manage the future workloads on HPC systems. However, there is a lack of appropriate models and frameworks to enable development, testing, and validation of new scheduling ideas.

In this paper, we present an open-source scheduler simulation framework (ScSF) that covers all the steps of scheduling research through simulation. ScSF provides capabilities for workload modeling, workload generation, system simulation, comparative workload analysis, and experiment orchestration. The simulator is designed to be run over a distributed computing infrastructure enabling to test at scale. We describe in detail a use case of ScSF to develop new techniques to manage scientific workflows in a batch scheduler. In the use case, such technique was implemented in the framework scheduler. For evaluation purposes, 1728 experiments, equivalent to 33 years of simulated time, were run in a deployment of ScSF over a distributed infrastructure of 17 compute nodes during two months. Finally, the experimental results were analyzed in the framework to judge that the technique minimizes workflows’ turnaround time without over-allocating resources. Finally, we discuss lessons learned from our experiences that will help future researchers.

Place, publisher, year, edition, pages
Springer, 2018
Series
Lecture Notes in Computer Science, ISSN 0302-9743, E-ISSN 1611-3349
Keywords
slurm, simulation, scheduling, HPC, High Performance Computing, workload, generation, analysis
National Category
Computer Sciences
Research subject
Computer Science
Identifiers
urn:nbn:se:umu:diva-132981 (URN)10.1007/978-3-319-77398-8_9 (DOI)000444863700009 ()978-3-319-77397-1 (ISBN)978-3-319-77398-8 (ISBN)
Conference
21th Workshop on Job Scheduling Strategies for Parallel Processing (JSSP 2017), Orlando FL, USA, June 2nd, 2017
Funder
eSSENCE - An eScience CollaborationSwedish Research Council, C0590801
Note

Work also supported by the U.S. Department of Energy, Office of Science, Office of Advanced Scientific Computing Research (ASCR) and we used resources at the National Energy Research Scientific Computing Center, a DOE Office of Science User Facility, supported by the Officece of Science of the U.S. Department of Energy, both under Contract No. DE-AC02-05CH11231.

Available from: 2017-03-27 Created: 2017-03-27 Last updated: 2018-10-05Bibliographically approved
Organisations

Search in DiVA

Show all publications