Today the cloud-desktop service, or Desktop-as-a-Service (DaaS), is massively replacing Virtual Desktop Infrastructures (VDI), as confirmed by the importance of players entering the DaaS market. In this paper we study the workload of a DaaS provider, analyzing three months of real traffic and resource usage. What emerges from the study, the first on the subject at the best of our knowledge, is that the workload on CPU and disk usage are long-tail distributed (lognormal, weibull and pare to) and that the length of working sessions is exponentially distributed. These results are extremely important for: the selection of the appropriate performance model to be used in capacity planning or run-time resource provisioning, the setup of workload generators, and the definition of heuristic policies for resource provisioning. The paper provides an accurate distribution fitting for all the workload features considered and discusses the implications of results on performance analysis.
{"title":"Cloud Desktop Workload: A Characterization Study","authors":"E. Casalicchio, Stefano Iannucci, L. Silvestri","doi":"10.1109/IC2E.2015.25","DOIUrl":"https://doi.org/10.1109/IC2E.2015.25","url":null,"abstract":"Today the cloud-desktop service, or Desktop-as-a-Service (DaaS), is massively replacing Virtual Desktop Infrastructures (VDI), as confirmed by the importance of players entering the DaaS market. In this paper we study the workload of a DaaS provider, analyzing three months of real traffic and resource usage. What emerges from the study, the first on the subject at the best of our knowledge, is that the workload on CPU and disk usage are long-tail distributed (lognormal, weibull and pare to) and that the length of working sessions is exponentially distributed. These results are extremely important for: the selection of the appropriate performance model to be used in capacity planning or run-time resource provisioning, the setup of workload generators, and the definition of heuristic policies for resource provisioning. The paper provides an accurate distribution fitting for all the workload features considered and discusses the implications of results on performance analysis.","PeriodicalId":395715,"journal":{"name":"2015 IEEE International Conference on Cloud Engineering","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130852867","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}
Privacy Level Agreements (PLAs) are likely to be increasingly adopted as a standardized way for cloud providers to describe their data protection practices. In this paper we propose an ontology-based model to represent the information disclosed in the agreement to turn it into a means that allows software tools to use and further process that information for different purposes, including automated service offering discovery and comparison. A specific usage of the PLA ontology is presented, showing how to link high level policies to operational policies that are then enforced and monitored. Through this established link, cloud users gain greater assurance that what is expressed in such agreements is actually being met, and thereby can take this information into account when choosing cloud service providers. Furthermore, the created link can be used to enable policy enforcement tools to add semantics to the evidence they produce; this mainly takes the form of logs that are associated with the specific policy of which execution they provide evidence. Furthermore, the use of the ontology model allows a means of enabling interoperability among tools that are in charge of the enforcement and monitoring of possible violations to the terms of the agreement.
{"title":"Towards a Formalised Representation for the Technical Enforcement of Privacy Level Agreements","authors":"Michela D'Errico, Siani Pearson","doi":"10.1109/IC2E.2015.72","DOIUrl":"https://doi.org/10.1109/IC2E.2015.72","url":null,"abstract":"Privacy Level Agreements (PLAs) are likely to be increasingly adopted as a standardized way for cloud providers to describe their data protection practices. In this paper we propose an ontology-based model to represent the information disclosed in the agreement to turn it into a means that allows software tools to use and further process that information for different purposes, including automated service offering discovery and comparison. A specific usage of the PLA ontology is presented, showing how to link high level policies to operational policies that are then enforced and monitored. Through this established link, cloud users gain greater assurance that what is expressed in such agreements is actually being met, and thereby can take this information into account when choosing cloud service providers. Furthermore, the created link can be used to enable policy enforcement tools to add semantics to the evidence they produce; this mainly takes the form of logs that are associated with the specific policy of which execution they provide evidence. Furthermore, the use of the ontology model allows a means of enabling interoperability among tools that are in charge of the enforcement and monitoring of possible violations to the terms of the agreement.","PeriodicalId":395715,"journal":{"name":"2015 IEEE International Conference on Cloud Engineering","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128661236","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}
Modern analytics services require the analysis of large quantities of data derived from disparate geo-distributed sources. Further, the analytics requirements can be complex, with many applications requiring a combination of both real-time and historical analysis, resulting in complex tradeoffs between cost, performance, and information quality. While the traditional approach to analytics processing is to send all the data to a dedicated centralized location, an alternative approach would be to push all computing to the edge for in-situ processing. We argue that neither approach is optimal for modern analytics requirements. Instead, we examine complex tradeoffs driven by a large number of factors such as application, data, and resource characteristics. We present an empirical study using Planet Lab experiments with beacon data from Akamai's download analytics service. We explore key tradeoffs and their implications for the design of next-generation scalable wide-area analytics.
{"title":"Towards Optimizing Wide-Area Streaming Analytics","authors":"Benjamin Heintz, A. Chandra, R. Sitaraman","doi":"10.1109/IC2E.2015.53","DOIUrl":"https://doi.org/10.1109/IC2E.2015.53","url":null,"abstract":"Modern analytics services require the analysis of large quantities of data derived from disparate geo-distributed sources. Further, the analytics requirements can be complex, with many applications requiring a combination of both real-time and historical analysis, resulting in complex tradeoffs between cost, performance, and information quality. While the traditional approach to analytics processing is to send all the data to a dedicated centralized location, an alternative approach would be to push all computing to the edge for in-situ processing. We argue that neither approach is optimal for modern analytics requirements. Instead, we examine complex tradeoffs driven by a large number of factors such as application, data, and resource characteristics. We present an empirical study using Planet Lab experiments with beacon data from Akamai's download analytics service. We explore key tradeoffs and their implications for the design of next-generation scalable wide-area analytics.","PeriodicalId":395715,"journal":{"name":"2015 IEEE International Conference on Cloud Engineering","volume":"404 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115920404","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}
D. Bhamare, R. Jain, M. Samaka, Gabor Vaszkun, A. Erbad
Network Function Virtualization (NFV) and Service Chaining (SC) are novel service deployment approaches in the contemporary cloud environments for increased flexibility and cost efficiency to the Application Service Providers and Network Providers. However, NFV and SC are still new and evolving topics. Optimized placement of these virtual functions is necessary for acceptable latency to the end-users. In this work we consider the problem of optimal Virtual Function (VF) placement in a multi-cloud environment to satisfy the client demands so that the total response time is minimized. In addition we consider the problem of dynamic service deployment for OpenADN, a novel multi-cloud application delivery platform.
{"title":"Multi-cloud Distribution of Virtual Functions and Dynamic Service Deployment: Open ADN Perspective","authors":"D. Bhamare, R. Jain, M. Samaka, Gabor Vaszkun, A. Erbad","doi":"10.1109/IC2E.2015.49","DOIUrl":"https://doi.org/10.1109/IC2E.2015.49","url":null,"abstract":"Network Function Virtualization (NFV) and Service Chaining (SC) are novel service deployment approaches in the contemporary cloud environments for increased flexibility and cost efficiency to the Application Service Providers and Network Providers. However, NFV and SC are still new and evolving topics. Optimized placement of these virtual functions is necessary for acceptable latency to the end-users. In this work we consider the problem of optimal Virtual Function (VF) placement in a multi-cloud environment to satisfy the client demands so that the total response time is minimized. In addition we consider the problem of dynamic service deployment for OpenADN, a novel multi-cloud application delivery platform.","PeriodicalId":395715,"journal":{"name":"2015 IEEE International Conference on Cloud Engineering","volume":"72 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116302305","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}
Infrastructure as a Service (IaaS) generally provides a standard vanilla server that contains an OS and basic functions, and each user has to manually install the required applications for the proper server deployments. We are working on a composite application deployment approach to automatically install selected applications in a flexible manner, based on a set of application installation scripts that are invoked on the vanilla server. Some applications have installation dependencies involving multiple servers. Previous research projects on installing applications with multi-server dependencies have deployed the servers sequentially. This means the total deployment time grows linearly with the number of servers. Our automated parallel approach makes the composite application deployment run in parallel when there are installation dependencies across multiple servers. We implemented a prototype system on Chef, a widely used automatic server installation framework, and evaluated the performance of our composite application deployment on a Soft Layer public cloud using two composite application server cases. The deployment times were reduced by roughly 40% in our trials.
{"title":"An Automated Parallel Approach for Rapid Deployment of Composite Application Servers","authors":"Yasuharu Katsuno, Hitomi Takahashi","doi":"10.1109/IC2E.2015.16","DOIUrl":"https://doi.org/10.1109/IC2E.2015.16","url":null,"abstract":"Infrastructure as a Service (IaaS) generally provides a standard vanilla server that contains an OS and basic functions, and each user has to manually install the required applications for the proper server deployments. We are working on a composite application deployment approach to automatically install selected applications in a flexible manner, based on a set of application installation scripts that are invoked on the vanilla server. Some applications have installation dependencies involving multiple servers. Previous research projects on installing applications with multi-server dependencies have deployed the servers sequentially. This means the total deployment time grows linearly with the number of servers. Our automated parallel approach makes the composite application deployment run in parallel when there are installation dependencies across multiple servers. We implemented a prototype system on Chef, a widely used automatic server installation framework, and evaluated the performance of our composite application deployment on a Soft Layer public cloud using two composite application server cases. The deployment times were reduced by roughly 40% in our trials.","PeriodicalId":395715,"journal":{"name":"2015 IEEE International Conference on Cloud Engineering","volume":"76 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127590172","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}
The spread of computing Clouds and block-chain technology enable a trend towards highly decentralized and distributed management of potentially very large datasets. Existing big data-mining systems are not designed for this nextlevel scale of distribution and decentralization. Additionally, the required system scalability is currently not given for very large datasets. We aim at solving these problems with a scalable distributed multi-agent system based approach.
{"title":"Multi-agent Based Intelligence Generation from Very Large Datasets","authors":"Karima Qayumi","doi":"10.1109/IC2E.2015.96","DOIUrl":"https://doi.org/10.1109/IC2E.2015.96","url":null,"abstract":"The spread of computing Clouds and block-chain technology enable a trend towards highly decentralized and distributed management of potentially very large datasets. Existing big data-mining systems are not designed for this nextlevel scale of distribution and decentralization. Additionally, the required system scalability is currently not given for very large datasets. We aim at solving these problems with a scalable distributed multi-agent system based approach.","PeriodicalId":395715,"journal":{"name":"2015 IEEE International Conference on Cloud Engineering","volume":"77 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134326534","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}
Over the last few years, more and more Cloud Computing offerings have emerged ranging from compute, data storage, and middleware services over platform environments up to ready-to-use applications. Choosing the best offering for a particular use case, is a complex task which involves comparison and trade-off analysis of functional and non-functional service properties; for non-functional quality of service (QoS) properties, this is typically done via benchmarking. Today, a plethora of benchmarking solutions exist for different layers in the cloud stack (IaaS, PaaS, SaaS) which typically address a single QoS dimension - a holistic cloud benchmark even for a single layer in the cloud stack is still missing. In this tutorial, we will give an overview of existing cloud benchmarking solutions and point-out ways in which these different benchmarks could be used in concert to actually compare clouds as a whole (i.e., for instance Amazon cloud vs. Google cloud) instead of analyzing isolated QoS dimensions of single cloud services.
{"title":"An Introduction to Cloud Benchmarking","authors":"David Bermbach","doi":"10.1109/IC2E.2015.65","DOIUrl":"https://doi.org/10.1109/IC2E.2015.65","url":null,"abstract":"Over the last few years, more and more Cloud Computing offerings have emerged ranging from compute, data storage, and middleware services over platform environments up to ready-to-use applications. Choosing the best offering for a particular use case, is a complex task which involves comparison and trade-off analysis of functional and non-functional service properties; for non-functional quality of service (QoS) properties, this is typically done via benchmarking. Today, a plethora of benchmarking solutions exist for different layers in the cloud stack (IaaS, PaaS, SaaS) which typically address a single QoS dimension - a holistic cloud benchmark even for a single layer in the cloud stack is still missing. In this tutorial, we will give an overview of existing cloud benchmarking solutions and point-out ways in which these different benchmarks could be used in concert to actually compare clouds as a whole (i.e., for instance Amazon cloud vs. Google cloud) instead of analyzing isolated QoS dimensions of single cloud services.","PeriodicalId":395715,"journal":{"name":"2015 IEEE International Conference on Cloud Engineering","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124555472","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}
In this work, we present a novel approach that leverages Linux containers to achieve High Availability (HA) for cloud applications. A middleware that is comprised of a set of HA agents is defined to compensate the limitations of Linux containers in achieving HA. In our approach we start modeling at the application level, considering the dependencies among application components. We generate the proper scheduling scheme and then deploy the application across containers in the cloud. For each container that hosts critical component(s), we continuously monitor its status and checkpoint its full state, and then react to its failure by restarting locally or failing over to another host where we resume the computing from the most recent state. By using this strategy, all components hosted in a container are preserved without intrusively imposing modification on the application side. Finally, the feasibility of our approach is verified by building a proof-of-concept prototype and a case study of a video streaming application.
{"title":"Leveraging Linux Containers to Achieve High Availability for Cloud Services","authors":"Wubin Li, A. Kanso, Abdelouahed Gherbi","doi":"10.1109/IC2E.2015.17","DOIUrl":"https://doi.org/10.1109/IC2E.2015.17","url":null,"abstract":"In this work, we present a novel approach that leverages Linux containers to achieve High Availability (HA) for cloud applications. A middleware that is comprised of a set of HA agents is defined to compensate the limitations of Linux containers in achieving HA. In our approach we start modeling at the application level, considering the dependencies among application components. We generate the proper scheduling scheme and then deploy the application across containers in the cloud. For each container that hosts critical component(s), we continuously monitor its status and checkpoint its full state, and then react to its failure by restarting locally or failing over to another host where we resume the computing from the most recent state. By using this strategy, all components hosted in a container are preserved without intrusively imposing modification on the application side. Finally, the feasibility of our approach is verified by building a proof-of-concept prototype and a case study of a video streaming application.","PeriodicalId":395715,"journal":{"name":"2015 IEEE International Conference on Cloud Engineering","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127734445","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}
The main quest for cloud stakeholders is to find an optimal deployment architecture for cloud applications that maximizes availability, minimizes cost, and addresses portability and scalability. Unfortunately, the lack of a unified definition and adequate modeling language and methodologies that address the cloud domain specific characteristics makes architecting efficient cloud applications a daunting task. This paper introduces Stratus ML: a technology agnostic integrated modeling framework for cloud applications. Stratus ML provides an intuitive user interface that allows the cloud stakeholders (i.e., providers, developers, administrators, and financial decision makers) to define their application services, configure them, specify the applications' behaviour at runtime through a set of adaptation rules, and estimate cost under diverse cloud platforms and configurations. Moreover, through a set of model transformation templates, Stratus ML maintains consistency between the various artifacts of cloud applications. This paper presents Stratus ML and illustrates its usefulness and practical applicability from different stakeholder perspectives. A demo video, usage scenario and other relevant information can be found at the Stratus ML webpage.
{"title":"Stratus ML: A Layered Cloud Modeling Framework","authors":"Mohammad Hamdaqa, L. Tahvildari","doi":"10.1109/IC2E.2015.42","DOIUrl":"https://doi.org/10.1109/IC2E.2015.42","url":null,"abstract":"The main quest for cloud stakeholders is to find an optimal deployment architecture for cloud applications that maximizes availability, minimizes cost, and addresses portability and scalability. Unfortunately, the lack of a unified definition and adequate modeling language and methodologies that address the cloud domain specific characteristics makes architecting efficient cloud applications a daunting task. This paper introduces Stratus ML: a technology agnostic integrated modeling framework for cloud applications. Stratus ML provides an intuitive user interface that allows the cloud stakeholders (i.e., providers, developers, administrators, and financial decision makers) to define their application services, configure them, specify the applications' behaviour at runtime through a set of adaptation rules, and estimate cost under diverse cloud platforms and configurations. Moreover, through a set of model transformation templates, Stratus ML maintains consistency between the various artifacts of cloud applications. This paper presents Stratus ML and illustrates its usefulness and practical applicability from different stakeholder perspectives. A demo video, usage scenario and other relevant information can be found at the Stratus ML webpage.","PeriodicalId":395715,"journal":{"name":"2015 IEEE International Conference on Cloud Engineering","volume":"375 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117083790","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}
Current data analytics software stacks are tailored to use large number of commodity machines in clusters, with each machine containing a small amount of memory. Thus, significant effort is made in these stacks to partition the data into small chunks, and process these chunks in parallel. Recent advances in memory technology now promise the availability of machines with the amount of memory increased by two or more orders of magnitude. For example, The Machine [1] currently under development at HP Labs plans to use memristor, a new type of non-volatile random access memory with much larger memory density at access speed comparable to today's dynamic random access memory. Such technologies offer the possibility of a flat memory/storage hierarchy, in-memory data processing and instant persistence of intermediate and final processing results. Photonic fabrics provide large communication bandwidth to move large volume of data between processing units at very low latency. Moreover, the multicore architectures adopt system-on-chip (SoC) designs to achieve significant compute performance with high power-efficiency.
{"title":"In-memory computing for scalable data analytics","authors":"Jun Yu Li","doi":"10.1109/IC2E.2015.59","DOIUrl":"https://doi.org/10.1109/IC2E.2015.59","url":null,"abstract":"Current data analytics software stacks are tailored to use large number of commodity machines in clusters, with each machine containing a small amount of memory. Thus, significant effort is made in these stacks to partition the data into small chunks, and process these chunks in parallel. Recent advances in memory technology now promise the availability of machines with the amount of memory increased by two or more orders of magnitude. For example, The Machine [1] currently under development at HP Labs plans to use memristor, a new type of non-volatile random access memory with much larger memory density at access speed comparable to today's dynamic random access memory. Such technologies offer the possibility of a flat memory/storage hierarchy, in-memory data processing and instant persistence of intermediate and final processing results. Photonic fabrics provide large communication bandwidth to move large volume of data between processing units at very low latency. Moreover, the multicore architectures adopt system-on-chip (SoC) designs to achieve significant compute performance with high power-efficiency.","PeriodicalId":395715,"journal":{"name":"2015 IEEE International Conference on Cloud Engineering","volume":"193 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114180801","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：