Pub Date : 2018-09-01DOI: 10.1109/FAS-W.2018.00030
A. Harutyunyan, A. Poghosyan, Naira Grigoryan, N. Kushmerick, Harutyun Beybutyan
The identification of important changes in a complex distributed system is a challenging data science problem. Solving this problem is critical for tools for managing modern cloud infrastructure stacks and other large complex distributed systems. In this paper, we investigate two specific approaches to using log data to solve this problem. The first approach is comparing a source's current and past behavior. Some solutions that perform anomaly detection on numeric data from the data center are inevitably relying on global change point detection concepts. On the other hand, while log data promises a significantly different perspectives and dimensions to accomplish a similar task, state-of-the-art of solutions lack a capability to automatically detect significant change points in the log stream of an event source through learning its behavioral patterns. Such change points indicate the most important times when the source's behavior significantly differs from the past. A second complementary approach to real-time change detection involves comparing a source's current behavior with the current behavior of its peers in a population of sources serving a common role in the data center. Employing the concept of event types of log messages introduced earlier, we propose algorithms for each of these approaches that apply classical statistical and machine learning techniques to data capturing the distribution of those constructs. We demonstrate experimental results from our prototype algorithms.
{"title":"Identifying Changed or Sick Resources from Logs","authors":"A. Harutyunyan, A. Poghosyan, Naira Grigoryan, N. Kushmerick, Harutyun Beybutyan","doi":"10.1109/FAS-W.2018.00030","DOIUrl":"https://doi.org/10.1109/FAS-W.2018.00030","url":null,"abstract":"The identification of important changes in a complex distributed system is a challenging data science problem. Solving this problem is critical for tools for managing modern cloud infrastructure stacks and other large complex distributed systems. In this paper, we investigate two specific approaches to using log data to solve this problem. The first approach is comparing a source's current and past behavior. Some solutions that perform anomaly detection on numeric data from the data center are inevitably relying on global change point detection concepts. On the other hand, while log data promises a significantly different perspectives and dimensions to accomplish a similar task, state-of-the-art of solutions lack a capability to automatically detect significant change points in the log stream of an event source through learning its behavioral patterns. Such change points indicate the most important times when the source's behavior significantly differs from the past. A second complementary approach to real-time change detection involves comparing a source's current behavior with the current behavior of its peers in a population of sources serving a common role in the data center. Employing the concept of event types of log messages introduced earlier, we propose algorithms for each of these approaches that apply classical statistical and machine learning techniques to data capturing the distribution of those constructs. We demonstrate experimental results from our prototype algorithms.","PeriodicalId":164903,"journal":{"name":"2018 IEEE 3rd International Workshops on Foundations and Applications of Self* Systems (FAS*W)","volume":"301 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123470564","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}
Pub Date : 2018-09-01DOI: 10.1109/FAS-W.2018.00035
Mirko D'Angelo, M. Caporuscio
Pure Edge Computing relies on peer-to-peer overlay networks to realize the communication backbone between participating entities. In these settings, entities are characterized by high heterogeneity, mobility, and variability, which introduce runtime uncertainty and may harm the dependability of the network. Departing from state-of-the-art solutions, overlay networks for Pure Edge Computing should take into account the dynamics of the operating environment and self-adapt their topology accordingly, in order to increase the dependability of the communication. To this end, this paper discusses the preliminary development and validation of SA-Chord, a self-adaptive version of the wellknown Chord protocol, able to adapt the network topology according to a given global goal. SA-Chord has been validated through simulation against two distinct goals: (i) minimize energy consumption and, (ii) maximize network throughput. Simulation results are promising and show how SA-Chord efficiently and effectively achieves a given goal.
{"title":"SA-Chord: A Self-Adaptive P2P Overlay Network","authors":"Mirko D'Angelo, M. Caporuscio","doi":"10.1109/FAS-W.2018.00035","DOIUrl":"https://doi.org/10.1109/FAS-W.2018.00035","url":null,"abstract":"Pure Edge Computing relies on peer-to-peer overlay networks to realize the communication backbone between participating entities. In these settings, entities are characterized by high heterogeneity, mobility, and variability, which introduce runtime uncertainty and may harm the dependability of the network. Departing from state-of-the-art solutions, overlay networks for Pure Edge Computing should take into account the dynamics of the operating environment and self-adapt their topology accordingly, in order to increase the dependability of the communication. To this end, this paper discusses the preliminary development and validation of SA-Chord, a self-adaptive version of the wellknown Chord protocol, able to adapt the network topology according to a given global goal. SA-Chord has been validated through simulation against two distinct goals: (i) minimize energy consumption and, (ii) maximize network throughput. Simulation results are promising and show how SA-Chord efficiently and effectively achieves a given goal.","PeriodicalId":164903,"journal":{"name":"2018 IEEE 3rd International Workshops on Foundations and Applications of Self* Systems (FAS*W)","volume":"104 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121873613","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}
Pub Date : 2018-09-01DOI: 10.1109/FAS-W.2018.00049
Sharmin Jahan, Charles Walter, Sarra M. Alqahtani, R. Gamble
Autonomous systems have become incredibly common, with autonomous vehicles and drones dictating major research trends. Coordination of autonomous vehicles is one of these trends. With multiple different, likely proprietary, systems all needing to communicate and accomplish a task as a unit, there is a need for each individual autonomous system to be capable of entering or leaving the unit, either because of a failure or the need to perform a different task. Thus, each device has a local goal it is trying to complete and a global goal that needs to be completed as part of the unit. Given environmental changes, the systems must adapt by determining how they can satisfy their local goals and self-integrate into the unit's goal when needed or when it is consistent with a local goal. In this paper, we examine self-integrating policies as part of satisfying a global goal when local goals also reside in an autonomous system. We use a Partial-Order, Causal-Link representation of a simple mission to discover potential flaws, or inconsistencies, present between two autonomous devices that affect the global mission. We use these flaws as triggers for self-integration. Assurance cases provide the medium to specify and validate the global and local mission constraints initially and upon adaptation. We demonstrate our solution using multiple Anki Cozmo robots to complete a multi-cube retrieval mission.
{"title":"Adaptive Coordination to Complete Mission Goals","authors":"Sharmin Jahan, Charles Walter, Sarra M. Alqahtani, R. Gamble","doi":"10.1109/FAS-W.2018.00049","DOIUrl":"https://doi.org/10.1109/FAS-W.2018.00049","url":null,"abstract":"Autonomous systems have become incredibly common, with autonomous vehicles and drones dictating major research trends. Coordination of autonomous vehicles is one of these trends. With multiple different, likely proprietary, systems all needing to communicate and accomplish a task as a unit, there is a need for each individual autonomous system to be capable of entering or leaving the unit, either because of a failure or the need to perform a different task. Thus, each device has a local goal it is trying to complete and a global goal that needs to be completed as part of the unit. Given environmental changes, the systems must adapt by determining how they can satisfy their local goals and self-integrate into the unit's goal when needed or when it is consistent with a local goal. In this paper, we examine self-integrating policies as part of satisfying a global goal when local goals also reside in an autonomous system. We use a Partial-Order, Causal-Link representation of a simple mission to discover potential flaws, or inconsistencies, present between two autonomous devices that affect the global mission. We use these flaws as triggers for self-integration. Assurance cases provide the medium to specify and validate the global and local mission constraints initially and upon adaptation. We demonstrate our solution using multiple Anki Cozmo robots to complete a multi-cube retrieval mission.","PeriodicalId":164903,"journal":{"name":"2018 IEEE 3rd International Workshops on Foundations and Applications of Self* Systems (FAS*W)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123531873","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}
Pub Date : 2018-09-01DOI: 10.1109/FAS-W.2018.00019
K. Eledlebi, D. Ruta, F. Saffre, Yousof Al-Hammadi, A. Isakovic
Efficient deployment of wireless sensor network (WSN) is one of the key challenges of the Internet of Things (IoT), and one where self-organizing processes and adaptation to obstacle-rich environments are critical. We developed a Voronoi tessellation based algorithm, BISON (Bio-inspired Self-organized Network), designed to insert and self-deploy nodes of WSN into any unknown, obstacle rich indoor environment, satisfying both, the coverage and the connectivity demands. To limit the power consumption and simulate realistic real-time environment discovery, BISON confines each node to use only locally sensed information, while avoiding obstacles and connecting with neighboring nodes. The algorithm is assessed in terms of the critical deployment evaluation metrics: the area coverage and distance traveled. The results reveal fast convergence to a fully connected network with low deployment costs.
{"title":"Voronoi-Based Indoor Deployment of Mobile Sensors Network with Obstacles","authors":"K. Eledlebi, D. Ruta, F. Saffre, Yousof Al-Hammadi, A. Isakovic","doi":"10.1109/FAS-W.2018.00019","DOIUrl":"https://doi.org/10.1109/FAS-W.2018.00019","url":null,"abstract":"Efficient deployment of wireless sensor network (WSN) is one of the key challenges of the Internet of Things (IoT), and one where self-organizing processes and adaptation to obstacle-rich environments are critical. We developed a Voronoi tessellation based algorithm, BISON (Bio-inspired Self-organized Network), designed to insert and self-deploy nodes of WSN into any unknown, obstacle rich indoor environment, satisfying both, the coverage and the connectivity demands. To limit the power consumption and simulate realistic real-time environment discovery, BISON confines each node to use only locally sensed information, while avoiding obstacles and connecting with neighboring nodes. The algorithm is assessed in terms of the critical deployment evaluation metrics: the area coverage and distance traveled. The results reveal fast convergence to a fully connected network with low deployment costs.","PeriodicalId":164903,"journal":{"name":"2018 IEEE 3rd International Workshops on Foundations and Applications of Self* Systems (FAS*W)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124078164","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}
Pub Date : 2018-09-01DOI: 10.1109/FAS-W.2018.00008
Leonel Aguilar, V. Bilano, Evangelos Pournaras, W. Maass, Pradeep Ravikumar, Julia Pueschel, C. Djeffal, S. Janzen, Giulio Rossetti, Stef Janssen, Fragkiskos D. Malliaros, L. Pappalardo, S. Ruggieri, Florin Pop, Josef Spillner, Johannes Klinglmayr, C. Leordeanu, Spyros Voulgaris, Takuto Sakamoto, Alexandra Carpen-Amarie
{"title":"DSS 2018 Foreword","authors":"Leonel Aguilar, V. Bilano, Evangelos Pournaras, W. Maass, Pradeep Ravikumar, Julia Pueschel, C. Djeffal, S. Janzen, Giulio Rossetti, Stef Janssen, Fragkiskos D. Malliaros, L. Pappalardo, S. Ruggieri, Florin Pop, Josef Spillner, Johannes Klinglmayr, C. Leordeanu, Spyros Voulgaris, Takuto Sakamoto, Alexandra Carpen-Amarie","doi":"10.1109/FAS-W.2018.00008","DOIUrl":"https://doi.org/10.1109/FAS-W.2018.00008","url":null,"abstract":"","PeriodicalId":164903,"journal":{"name":"2018 IEEE 3rd International Workshops on Foundations and Applications of Self* Systems (FAS*W)","volume":"67 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126213222","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}
Pub Date : 2018-09-01DOI: 10.1109/FAS-W.2018.00036
L. Sabatucci, M. Cossentino, G. D. Simone, S. Lopes
The functioning of the Shipboard Power System (SPS) is critical to the survival and safety of the ship because many accidents occurring during ship navigation are often due to electrical failures. In smart vessels, the SPS reconfiguration consists of a variation of the electrical topology to successfully supply energy to critical services. The proposed reconfiguration procedure uses a distributed and mission-oriented approach, and it employs a generic-purpose self-adaptive middleware (MUSA). MUSA has been customized to dynamically reconfigure an SPS in case of failures or unexpected events. It allows obtaining a runtime solution that properly considers ships mission and current scenario. We also implemented an experimental setup including a Matlab/Simulink simulation of a case study from literature.
{"title":"Self-Reconfiguration of Shipboard Power Systems","authors":"L. Sabatucci, M. Cossentino, G. D. Simone, S. Lopes","doi":"10.1109/FAS-W.2018.00036","DOIUrl":"https://doi.org/10.1109/FAS-W.2018.00036","url":null,"abstract":"The functioning of the Shipboard Power System (SPS) is critical to the survival and safety of the ship because many accidents occurring during ship navigation are often due to electrical failures. In smart vessels, the SPS reconfiguration consists of a variation of the electrical topology to successfully supply energy to critical services. The proposed reconfiguration procedure uses a distributed and mission-oriented approach, and it employs a generic-purpose self-adaptive middleware (MUSA). MUSA has been customized to dynamically reconfigure an SPS in case of failures or unexpected events. It allows obtaining a runtime solution that properly considers ships mission and current scenario. We also implemented an experimental setup including a Matlab/Simulink simulation of a case study from literature.","PeriodicalId":164903,"journal":{"name":"2018 IEEE 3rd International Workshops on Foundations and Applications of Self* Systems (FAS*W)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132492332","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}
Pub Date : 2018-09-01DOI: 10.1109/FAS-W.2018.00051
C. Landauer
When we build systems to operate in hazardous or remote environments, and especially when we expect them to cooperate with others in support of a goal, we rely on them to operate as correctly as possible in the (unpredicted, frequently unpredictable) situations they encounter. But the environment does whatever it does; we have essentially no control and only limited knowledge of what it does, and only the most meager notion of what it will do. In this paper, we describe an architecture for these component systems that we think will be better suited to the vagaries of environmental behavior than others. We advocate a collection of subsidiary systems to operate in parallel with the main system, to act as "outriggers" for unexpected environmental behaviors or system failures, or as "training wheels" during development. We describe our initial notions of how they relate to the original system and how to implement them.
{"title":"Outriggers and Training Wheels for Cooperating Systems","authors":"C. Landauer","doi":"10.1109/FAS-W.2018.00051","DOIUrl":"https://doi.org/10.1109/FAS-W.2018.00051","url":null,"abstract":"When we build systems to operate in hazardous or remote environments, and especially when we expect them to cooperate with others in support of a goal, we rely on them to operate as correctly as possible in the (unpredicted, frequently unpredictable) situations they encounter. But the environment does whatever it does; we have essentially no control and only limited knowledge of what it does, and only the most meager notion of what it will do. In this paper, we describe an architecture for these component systems that we think will be better suited to the vagaries of environmental behavior than others. We advocate a collection of subsidiary systems to operate in parallel with the main system, to act as \"outriggers\" for unexpected environmental behaviors or system failures, or as \"training wheels\" during development. We describe our initial notions of how they relate to the original system and how to implement them.","PeriodicalId":164903,"journal":{"name":"2018 IEEE 3rd International Workshops on Foundations and Applications of Self* Systems (FAS*W)","volume":"61 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132185092","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}
Pub Date : 2018-09-01DOI: 10.1109/FAS-W.2018.00010
Heiko Hamann, S. Mammen, Ingo Mauser, P. Ayres, Wolfgang Banzhaff, P. Bentley, P. Dittrich, M. Dorigo, R. Doursat, J. Hensen, W. Höhl, C. Jacob, A. Menges, Olivier Michel, Niels Napp, Kirstin H. Petersen, Hiroki Sayama, T. Schmickl, K. Støy, G. Theraulaz, Justin Werfel, A. Zamuda
{"title":"SOCO 2018 Foreword","authors":"Heiko Hamann, S. Mammen, Ingo Mauser, P. Ayres, Wolfgang Banzhaff, P. Bentley, P. Dittrich, M. Dorigo, R. Doursat, J. Hensen, W. Höhl, C. Jacob, A. Menges, Olivier Michel, Niels Napp, Kirstin H. Petersen, Hiroki Sayama, T. Schmickl, K. Støy, G. Theraulaz, Justin Werfel, A. Zamuda","doi":"10.1109/FAS-W.2018.00010","DOIUrl":"https://doi.org/10.1109/FAS-W.2018.00010","url":null,"abstract":"","PeriodicalId":164903,"journal":{"name":"2018 IEEE 3rd International Workshops on Foundations and Applications of Self* Systems (FAS*W)","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130603452","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}
Pub Date : 2018-09-01DOI: 10.1109/FAS-W.2018.00041
Yiwen Hua, Yawen Deng, Kirstin H. Petersen
The TERMES system is a robot collective capable of constructing 2.5D user-specified structures with specialized bricks. This work extends the original system, by enabling 3D construction without added complexity in the robots. To do this, we introduce an expandable brick which complies with the original TERMES hardware and is inexpensive and fast to fabricate. We further show a decentralized algorithm that permits an arbitrary number of robots to use both original and expandable bricks to build structures with overhangs over convex cavities, i.e. with bridges and roofs. Finally, we discuss a mechanical redesign of the robots towards decreased system cost, fabrication and maintenance time. Although more work is needed to realize construction of large-scale overhangs in practice, our work represents an important step towards construction of complex structures by minimalistic and scalable robot collectives.
{"title":"Robots Building Bridges, Not Walls","authors":"Yiwen Hua, Yawen Deng, Kirstin H. Petersen","doi":"10.1109/FAS-W.2018.00041","DOIUrl":"https://doi.org/10.1109/FAS-W.2018.00041","url":null,"abstract":"The TERMES system is a robot collective capable of constructing 2.5D user-specified structures with specialized bricks. This work extends the original system, by enabling 3D construction without added complexity in the robots. To do this, we introduce an expandable brick which complies with the original TERMES hardware and is inexpensive and fast to fabricate. We further show a decentralized algorithm that permits an arbitrary number of robots to use both original and expandable bricks to build structures with overhangs over convex cavities, i.e. with bridges and roofs. Finally, we discuss a mechanical redesign of the robots towards decreased system cost, fabrication and maintenance time. Although more work is needed to realize construction of large-scale overhangs in practice, our work represents an important step towards construction of complex structures by minimalistic and scalable robot collectives.","PeriodicalId":164903,"journal":{"name":"2018 IEEE 3rd International Workshops on Foundations and Applications of Self* Systems (FAS*W)","volume":"55 9","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120882609","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}
Pub Date : 2018-09-01DOI: 10.1109/FAS-W.2018.00028
Min Jeesoo, Sung Hanul, Eom Hyeonsang
As cloud data centers are dramatically growing, various applications are moved to cloud data centers owing to cost benefits for maintenance and hardware resources. However, latency-critical workloads among them suffer from some problems to fully achieve the cost effectiveness. The latency-critical workloads should show latencies in a stable manner, to be predicted, for strictly meeting QoSs. However, if they are executed with other workloads to save the cost, they experience QoS violation due to the contention for the hardware resources shared with co-location workloads. In order to guarantee QoSs and to improve the hardware resourse utilization, we proposed a memory bandwidth management method with an effective prediction model using machine learning. The prediction model estimates the amount of memory bandwidth that will be allocated to the latency-critical workload based on a REP decision tree. To construct this model, we first collect data and train the model with the data. The generated model can estimate the amount of memory bandwidth for meeting the SLO of the latency-critical workload no matter what batch processing workloads are collocated. The use of our approach achieves up to 99% SLO assurance and improves the server utilization up to 6.8x on average.
{"title":"OMBM-ML: An Efficient Memory Bandwidth Management for Ensuring QoS and Improving Server Utilization","authors":"Min Jeesoo, Sung Hanul, Eom Hyeonsang","doi":"10.1109/FAS-W.2018.00028","DOIUrl":"https://doi.org/10.1109/FAS-W.2018.00028","url":null,"abstract":"As cloud data centers are dramatically growing, various applications are moved to cloud data centers owing to cost benefits for maintenance and hardware resources. However, latency-critical workloads among them suffer from some problems to fully achieve the cost effectiveness. The latency-critical workloads should show latencies in a stable manner, to be predicted, for strictly meeting QoSs. However, if they are executed with other workloads to save the cost, they experience QoS violation due to the contention for the hardware resources shared with co-location workloads. In order to guarantee QoSs and to improve the hardware resourse utilization, we proposed a memory bandwidth management method with an effective prediction model using machine learning. The prediction model estimates the amount of memory bandwidth that will be allocated to the latency-critical workload based on a REP decision tree. To construct this model, we first collect data and train the model with the data. The generated model can estimate the amount of memory bandwidth for meeting the SLO of the latency-critical workload no matter what batch processing workloads are collocated. The use of our approach achieves up to 99% SLO assurance and improves the server utilization up to 6.8x on average.","PeriodicalId":164903,"journal":{"name":"2018 IEEE 3rd International Workshops on Foundations and Applications of Self* Systems (FAS*W)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130247325","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：