Pub Date : 2020-02-03DOI: 10.1504/ijguc.2020.10026548
Jia-Chun Lin, Ming-Chang Lee, Ingrid Chieh Yu, E. Johnsen
Streams of data are produced today at an unprecedented scale. Efficient and stable processing of these streams requires a careful interplay between the parameters of the streaming application and of the underlying stream processing framework. Today, finding these parameters happens by trial and error on the complex, deployed framework. This paper shows that high-level models can help to determine these parameters by predicting and comparing the performance of streaming applications running on stream processing frameworks with different configurations. To demonstrate this approach, this paper considers Spark Streaming, a widely used framework to leverage data streams on the fly and provide real-time stream processing. Technically, we develop a configurable and executable model to simulate both the streaming applications and the underlying Spark stream processing framework. Furthermore, we model the deployment of Spark Streaming on Apache YARN, which is a popular open-source distributed software framework for big data processing. We show that the developed model provides a satisfactory accuracy for predicting performance by means of empirical validation.
{"title":"A configurable and executable model of Spark Streaming on Apache YARN","authors":"Jia-Chun Lin, Ming-Chang Lee, Ingrid Chieh Yu, E. Johnsen","doi":"10.1504/ijguc.2020.10026548","DOIUrl":"https://doi.org/10.1504/ijguc.2020.10026548","url":null,"abstract":"Streams of data are produced today at an unprecedented scale. Efficient and stable processing of these streams requires a careful interplay between the parameters of the streaming application and of the underlying stream processing framework. Today, finding these parameters happens by trial and error on the complex, deployed framework. This paper shows that high-level models can help to determine these parameters by predicting and comparing the performance of streaming applications running on stream processing frameworks with different configurations. To demonstrate this approach, this paper considers Spark Streaming, a widely used framework to leverage data streams on the fly and provide real-time stream processing. Technically, we develop a configurable and executable model to simulate both the streaming applications and the underlying Spark stream processing framework. Furthermore, we model the deployment of Spark Streaming on Apache YARN, which is a popular open-source distributed software framework for big data processing. We show that the developed model provides a satisfactory accuracy for predicting performance by means of empirical validation.","PeriodicalId":375871,"journal":{"name":"Int. J. Grid Util. Comput.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131148395","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 : 2020-02-03DOI: 10.1504/ijguc.2020.10026557
R. Yaqub, F. A. Shifa, M. F. Butt
The proliferation of Electric Vehicles (EVs) is exponential. However, the power grid is not able to provide simultaneous charging of several EVs due to limited power production capabilities and old distribution infrastructure. Scheduled charging is one of the most advocated solutions. However, it is not viable for emergency vehicles. This paper proposes to provide priority charging service for government authorised emergency EVs. For enablement of this proposal, a complete solution that includes architecture, as well as the protocols suite is suggested. To realise such a service, the paper suggests a major functional entity called a Priority Charging Server, i.e. a database server where authorised emergency EVs' IDs are registered, and their record is maintained. The paper also proposes modifications in the IEC15118 and IEC 61850 protocol suites. These protocols provide communication between vehicle and grid. The solution also inculcates roaming as well as non-roaming scenarios, i.e. a priority charging request may be originated by an authorised emergency EV from a Home Utility Network, as well as Visiting Utility Network. The paper is concluded with a MATLAB-based proof-of-concept simulation.
{"title":"Preferential charging for government authorised emergency electric vehicles","authors":"R. Yaqub, F. A. Shifa, M. F. Butt","doi":"10.1504/ijguc.2020.10026557","DOIUrl":"https://doi.org/10.1504/ijguc.2020.10026557","url":null,"abstract":"The proliferation of Electric Vehicles (EVs) is exponential. However, the power grid is not able to provide simultaneous charging of several EVs due to limited power production capabilities and old distribution infrastructure. Scheduled charging is one of the most advocated solutions. However, it is not viable for emergency vehicles. This paper proposes to provide priority charging service for government authorised emergency EVs. For enablement of this proposal, a complete solution that includes architecture, as well as the protocols suite is suggested. To realise such a service, the paper suggests a major functional entity called a Priority Charging Server, i.e. a database server where authorised emergency EVs' IDs are registered, and their record is maintained. The paper also proposes modifications in the IEC15118 and IEC 61850 protocol suites. These protocols provide communication between vehicle and grid. The solution also inculcates roaming as well as non-roaming scenarios, i.e. a priority charging request may be originated by an authorised emergency EV from a Home Utility Network, as well as Visiting Utility Network. The paper is concluded with a MATLAB-based proof-of-concept simulation.","PeriodicalId":375871,"journal":{"name":"Int. J. Grid Util. Comput.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131296245","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 : 2020-02-03DOI: 10.1504/ijguc.2020.10026555
W. Osborn
In this paper, two strategies for handling continuous k-nearest neighbour queries for location-based services are proposed. CKNN1 and CKNN2 utilise a validity (i.e., safe) region approach for minimising the number of query requests that need to be send to the server. They also use a two-dimensional spatial access method for both validity region selection and in-structure searching. The latter feature ensures that new searches for a validity region are not required to begin from the root. An evaluation and comparison of both strategies is performed against repeated nearest neighbour search. Both random and exponentially distributed point sets are utilised. Results show that both approaches achieve significant performance gains, especially with respect to reducing the number of queries that must be sent from the client to the server.
{"title":"A spatial access method approach to continuous k-nearest neighbour processing for location-based services","authors":"W. Osborn","doi":"10.1504/ijguc.2020.10026555","DOIUrl":"https://doi.org/10.1504/ijguc.2020.10026555","url":null,"abstract":"In this paper, two strategies for handling continuous k-nearest neighbour queries for location-based services are proposed. CKNN1 and CKNN2 utilise a validity (i.e., safe) region approach for minimising the number of query requests that need to be send to the server. They also use a two-dimensional spatial access method for both validity region selection and in-structure searching. The latter feature ensures that new searches for a validity region are not required to begin from the root. An evaluation and comparison of both strategies is performed against repeated nearest neighbour search. Both random and exponentially distributed point sets are utilised. Results show that both approaches achieve significant performance gains, especially with respect to reducing the number of queries that must be sent from the client to the server.","PeriodicalId":375871,"journal":{"name":"Int. J. Grid Util. Comput.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122119742","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 : 2020-02-03DOI: 10.1504/ijguc.2020.10026550
Carlos Melo, J. Dantas, P. Maciel, Danilo Oliveira, J. Araujo, Rúbens de Souza Matos Júnior, Iure Fé
The Data Centre concept has evolved, mainly due to the need to reduce expenses with the required physical space to store, provide and maintain large computational infrastructures. The Software-Defined Data Centre (SDDC) is a result of this evolution. Through SDDC, any service can be hosted by virtualising more reliable and easier-to-keep hardware resources. Nowadays, many services and resources can be provided in a single rack, or even a single machine, with similar availability, considering the deployment costs of silo-based environments. One of the ways to apply the SDDC into a data centre is through hyper-convergence. Among the main contributions of this paper are the behavioural models developed for availability and capacity-oriented availability evaluation of silo-based, converged and hyper-converged cloud computing infrastructures. The obtained results may help stakeholders to select between converged and hyper-converged environments, due to their similar availability but with the particularity of having lower deployment costs.
{"title":"Models for hyper-converged cloud computing infrastructures planning","authors":"Carlos Melo, J. Dantas, P. Maciel, Danilo Oliveira, J. Araujo, Rúbens de Souza Matos Júnior, Iure Fé","doi":"10.1504/ijguc.2020.10026550","DOIUrl":"https://doi.org/10.1504/ijguc.2020.10026550","url":null,"abstract":"The Data Centre concept has evolved, mainly due to the need to reduce expenses with the required physical space to store, provide and maintain large computational infrastructures. The Software-Defined Data Centre (SDDC) is a result of this evolution. Through SDDC, any service can be hosted by virtualising more reliable and easier-to-keep hardware resources. Nowadays, many services and resources can be provided in a single rack, or even a single machine, with similar availability, considering the deployment costs of silo-based environments. One of the ways to apply the SDDC into a data centre is through hyper-convergence. Among the main contributions of this paper are the behavioural models developed for availability and capacity-oriented availability evaluation of silo-based, converged and hyper-converged cloud computing infrastructures. The obtained results may help stakeholders to select between converged and hyper-converged environments, due to their similar availability but with the particularity of having lower deployment costs.","PeriodicalId":375871,"journal":{"name":"Int. J. Grid Util. Comput.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127853063","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 : 2019-08-09DOI: 10.1504/IJGUC.2019.10023104
G. Morana, Rao V. Mikkilineni, Surendra Keshan
A new computing paradigm using Distributed Intelligent Managed Elements (DIME) and DIME Network Architecture (DNA) is used to demonstrate globally interoperable public and private cloud network deploying cloud agnostic workloads. The workloads are cognitive and able to adjust autonomously their structure and maintain desired quality-of-service (QoS). DNA is designed to provide a control architecture for workload self-management of non-functional requirements to address rapid fluctuations either in workload demand or fluctuations in available resources. Using DNA, a transaction intensive three-tier workload is migrated from a physical server to a virtual machine hosted in a public cloud without interrupting the service transactions. After migration, cloud agnostic inter-cloud and intra-cloud auto-scaling, auto- failover and live migration are demonstrated again, without disrupting the user experience or losing transactions.
{"title":"Cognitive workload management on globally interoperable network of clouds","authors":"G. Morana, Rao V. Mikkilineni, Surendra Keshan","doi":"10.1504/IJGUC.2019.10023104","DOIUrl":"https://doi.org/10.1504/IJGUC.2019.10023104","url":null,"abstract":"A new computing paradigm using Distributed Intelligent Managed Elements (DIME) and DIME Network Architecture (DNA) is used to demonstrate globally interoperable public and private cloud network deploying cloud agnostic workloads. The workloads are cognitive and able to adjust autonomously their structure and maintain desired quality-of-service (QoS). DNA is designed to provide a control architecture for workload self-management of non-functional requirements to address rapid fluctuations either in workload demand or fluctuations in available resources. Using DNA, a transaction intensive three-tier workload is migrated from a physical server to a virtual machine hosted in a public cloud without interrupting the service transactions. After migration, cloud agnostic inter-cloud and intra-cloud auto-scaling, auto- failover and live migration are demonstrated again, without disrupting the user experience or losing transactions.","PeriodicalId":375871,"journal":{"name":"Int. J. Grid Util. Comput.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2019-08-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115398922","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 : 2019-08-09DOI: 10.1504/IJGUC.2019.10023115
Chunfen Xu
This paper studies the linkages among the industries based on CSI 800 industry index, which provides mass complicated data for industry research. Factor analysis, a useful data analysis tool, allows researchers to investigate concepts that are not easily measured directly by collapsing a large number of variables into a few interpretable underlying factors. Firstly, data of ten industries in the period from September 2009 to March 2017 is collected from CSI 800 Index and correlational analyses are conducted. Secondly, this paper establishes an appropriate evaluation system, and then uses factor analysis to do dimension reduction. Finally, some characteristics and trends in various industries are obtained.
{"title":"Data analysis of CSI 800 industry index by using factor analysis model","authors":"Chunfen Xu","doi":"10.1504/IJGUC.2019.10023115","DOIUrl":"https://doi.org/10.1504/IJGUC.2019.10023115","url":null,"abstract":"This paper studies the linkages among the industries based on CSI 800 industry index, which provides mass complicated data for industry research. Factor analysis, a useful data analysis tool, allows researchers to investigate concepts that are not easily measured directly by collapsing a large number of variables into a few interpretable underlying factors. Firstly, data of ten industries in the period from September 2009 to March 2017 is collected from CSI 800 Index and correlational analyses are conducted. Secondly, this paper establishes an appropriate evaluation system, and then uses factor analysis to do dimension reduction. Finally, some characteristics and trends in various industries are obtained.","PeriodicalId":375871,"journal":{"name":"Int. J. Grid Util. Comput.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2019-08-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131391268","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 : 2019-08-09DOI: 10.1504/IJGUC.2019.10023134
A. S. Raugust, W. J. Loch, Felipe Rodrigo de Souza, M. A. Pillon, C. Miers, G. Koslovski
The diversity of services, prices, and geographical footprints has turned the clouds into a complex and heterogeneous environment. Moreover, the survivability and reliability aspects are often disregarded by tenants, eventually resulting in heavy losses due to unavailability of services that are hosted on Virtual Infrastructures (VIs). We present an alternative to improve Vis' survivability and reliability, which considers the use of replicas and the spreading of virtual resources atop providers, regions, and zones. We formulate the VI allocation with survivability and reliability requirements as a mixed integer program, and three strategies to solve the formulation are proposed. First, the binary constraints are relaxed to obtain a Linear Program (LP), and the LP solution is given as input for the simulated annealing technique. Then, two GPU-accelerated algorithms are proposed to speed up the allocation of large-scale scenarios. Simulation with different reliability requests indicate an increasing in survivability without inflating costs.
{"title":"Towards to virtual infrastructure allocation on multiple IaaS providers with survivability and reliability requirements","authors":"A. S. Raugust, W. J. Loch, Felipe Rodrigo de Souza, M. A. Pillon, C. Miers, G. Koslovski","doi":"10.1504/IJGUC.2019.10023134","DOIUrl":"https://doi.org/10.1504/IJGUC.2019.10023134","url":null,"abstract":"The diversity of services, prices, and geographical footprints has turned the clouds into a complex and heterogeneous environment. Moreover, the survivability and reliability aspects are often disregarded by tenants, eventually resulting in heavy losses due to unavailability of services that are hosted on Virtual Infrastructures (VIs). We present an alternative to improve Vis' survivability and reliability, which considers the use of replicas and the spreading of virtual resources atop providers, regions, and zones. We formulate the VI allocation with survivability and reliability requirements as a mixed integer program, and three strategies to solve the formulation are proposed. First, the binary constraints are relaxed to obtain a Linear Program (LP), and the LP solution is given as input for the simulated annealing technique. Then, two GPU-accelerated algorithms are proposed to speed up the allocation of large-scale scenarios. Simulation with different reliability requests indicate an increasing in survivability without inflating costs.","PeriodicalId":375871,"journal":{"name":"Int. J. Grid Util. Comput.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2019-08-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125602057","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 : 2019-08-09DOI: 10.1504/IJGUC.2019.102748
Mohamed Mehdi Kandi, Shaoyi Yin, A. Hameurlain
Cloud computing is a technology that provides on-demand services in which the number of assigned resources can be automatically adjusted. A key challenge is how to choose the right number of resources so that the overall monetary cost is minimised. This problem, known as auto-scaling, was addressed in some existing works but most of them are dedicated to web applications. In these applications, it is assumed that the queries are atomic and each of them uses a single resource for a short period of time. However, this assumption cannot be considered for database applications. A query, in this case, contains many dependent and long tasks so several resources are required. We propose in this work an auto-scaling method based on reinforcement learning. The method is coupled with placement-scheduling. In the experimental section, we show the advantage of coupling the auto-scaling to the placement-scheduling by comparing our work to an existing auto-scaling method.
{"title":"Resource auto-scaling for SQL-like queries in the cloud based on parallel reinforcement learning","authors":"Mohamed Mehdi Kandi, Shaoyi Yin, A. Hameurlain","doi":"10.1504/IJGUC.2019.102748","DOIUrl":"https://doi.org/10.1504/IJGUC.2019.102748","url":null,"abstract":"Cloud computing is a technology that provides on-demand services in which the number of assigned resources can be automatically adjusted. A key challenge is how to choose the right number of resources so that the overall monetary cost is minimised. This problem, known as auto-scaling, was addressed in some existing works but most of them are dedicated to web applications. In these applications, it is assumed that the queries are atomic and each of them uses a single resource for a short period of time. However, this assumption cannot be considered for database applications. A query, in this case, contains many dependent and long tasks so several resources are required. We propose in this work an auto-scaling method based on reinforcement learning. The method is coupled with placement-scheduling. In the experimental section, we show the advantage of coupling the auto-scaling to the placement-scheduling by comparing our work to an existing auto-scaling method.","PeriodicalId":375871,"journal":{"name":"Int. J. Grid Util. Comput.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2019-08-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116577931","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 : 2019-08-09DOI: 10.1504/IJGUC.2020.10021245
A. Longo, M. Zappatore, Mario Alessandro Bochicchio
Public contracts regulate how public services are managed by the stakeholders. However, the current technological trend is creating a significant bias between the pace at which service data are produced and that at which contracts change. This increased availability of service data can be exploited in public procurement processes by fostering novel approaches to manage contracts, making them more dynamic and improving the Quality of Service (QoS) delivered to customers. In this paper, a cloud-based approach for assessing the QoS in Local Transportation Services (LTSs) in Apulia Region (Southern Italy) is proposed. Service Level Agreements (SLAs) between providers and the Regional Authority, as well as the minimal guaranteed QoS levels between providers and passengers, are modelled as contracts enacted via a cloud-based system, which gathers data from sensors and passengers. In this way, changes in contract conditions for improving the perceived and delivered QoS can be fastened and facilitated based on data. In order to validate the pilot case, a set of quality indicators and service levels grounded in European and Italian regulatory frameworks has been considered.
{"title":"A cloud-based approach to dynamically manage service contracts for local public transportation","authors":"A. Longo, M. Zappatore, Mario Alessandro Bochicchio","doi":"10.1504/IJGUC.2020.10021245","DOIUrl":"https://doi.org/10.1504/IJGUC.2020.10021245","url":null,"abstract":"Public contracts regulate how public services are managed by the stakeholders. However, the current technological trend is creating a significant bias between the pace at which service data are produced and that at which contracts change. This increased availability of service data can be exploited in public procurement processes by fostering novel approaches to manage contracts, making them more dynamic and improving the Quality of Service (QoS) delivered to customers. In this paper, a cloud-based approach for assessing the QoS in Local Transportation Services (LTSs) in Apulia Region (Southern Italy) is proposed. Service Level Agreements (SLAs) between providers and the Regional Authority, as well as the minimal guaranteed QoS levels between providers and passengers, are modelled as contracts enacted via a cloud-based system, which gathers data from sensors and passengers. In this way, changes in contract conditions for improving the perceived and delivered QoS can be fastened and facilitated based on data. In order to validate the pilot case, a set of quality indicators and service levels grounded in European and Italian regulatory frameworks has been considered.","PeriodicalId":375871,"journal":{"name":"Int. J. Grid Util. Comput.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2019-08-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121221448","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 : 2019-08-09DOI: 10.1504/IJGUC.2019.10023108
Paul Moggridge, N. Helian, Yi Sun, M. Lilley, V. Veneziano, Martin Eaves
In this paper, the Max-Min Fast Track (MXFT) scheduling algorithm is improved and compared against a selection of popular algorithms. The improved versions of MXFT are called Min-Min Max-Min Fast Track (MMMXFT) and Clustering Min-Min Max-Min Fast Track (CMMMXFT). The key difference is using Min-Min for the fast track. Experimentation revealed that despite Min-Min's characteristic of prioritising small tasks at the expense of overall makespan, the overall makespan was not adversely affected and the benefits of prioritising small tasks were identified in MMMXFT. Experiments were conducted by using a simulator with the exception of one real-world experiment. The real-world experiment identified challenges faced by algorithms which rely on accurate execution time prediction.
本文对最大-最小快速通道调度算法进行了改进，并与一些常用算法进行了比较。MXFT的改进版本被称为Min-Min Max-Min Fast Track (MMMXFT)和Clustering Min-Min Max-Min Fast Track (CMMMXFT)。关键的区别是使用Min-Min快速通道。实验表明，尽管Min-Min以牺牲总体完工时间为代价优先考虑小任务的特点，但总体完工时间并没有受到不利影响，并且在MMMXFT中确定了优先考虑小任务的好处。除一次真实实验外，其余实验均使用模拟器进行。现实世界的实验表明，依赖于准确的执行时间预测的算法所面临的挑战。
{"title":"Improving the MXFT scheduling algorithm for a cloud computing context","authors":"Paul Moggridge, N. Helian, Yi Sun, M. Lilley, V. Veneziano, Martin Eaves","doi":"10.1504/IJGUC.2019.10023108","DOIUrl":"https://doi.org/10.1504/IJGUC.2019.10023108","url":null,"abstract":"In this paper, the Max-Min Fast Track (MXFT) scheduling algorithm is improved and compared against a selection of popular algorithms. The improved versions of MXFT are called Min-Min Max-Min Fast Track (MMMXFT) and Clustering Min-Min Max-Min Fast Track (CMMMXFT). The key difference is using Min-Min for the fast track. Experimentation revealed that despite Min-Min's characteristic of prioritising small tasks at the expense of overall makespan, the overall makespan was not adversely affected and the benefits of prioritising small tasks were identified in MMMXFT. Experiments were conducted by using a simulator with the exception of one real-world experiment. The real-world experiment identified challenges faced by algorithms which rely on accurate execution time prediction.","PeriodicalId":375871,"journal":{"name":"Int. J. Grid Util. Comput.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2019-08-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123111184","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：