Managing energy, performance and cost in large scale heterogeneous datacenters using migrations

Zakarya, Muhammad; Gillam, Lee

doi:10.1016/j.future.2018.10.044

Cited by 27 publications

(31 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As, there are two instances running during the migration duration (one on source host and a second one on the target host); the energy cost relates to the consumed energy by the least energy efficient host [11]. Moreover, in our evaluation, we account for 10% performance degradation due to migration that is merely modelled based on the memory, disk size and available network bandwidth [1], [23]; however, complete network heterogeneity is not within the scope of this paper. Similarly, co-locating 18 VMs when they compete for similar resources can also degrade VMs performance [31].…”

Section: Validation Using Experimental Results (Planetlab Dataset)mentioning

confidence: 99%

“…From an experimental point of view, when a particular application is being migrated from one host to another host, the impact of host heterogeneity is added as an increase or a decrease in application's runtime. Further details on these parameters and the mapping mechanism to model resource and application heterogeneities can be found in [11], [23]. Povray benchmarks on E5430, E5-2650 and E5645; where, E5430 performs better than E5645 [11] We extend the PowerHost class of the CloudSim simulator to account for platform heterogeneities.…”

Section: Modelling Resource and Application Heterogeneitiesmentioning

confidence: 99%

See 1 more Smart Citation

Modelling resource heterogeneities in cloud simulations and quantifying their accuracy

Zakarya

Gillam

2019

Simulation Modelling Practice and Theory

Self Cite

View full text Add to dashboard Cite

Section: Validation Using Experimental Results (Planetlab Dataset)mentioning

confidence: 99%

Section: Modelling Resource and Application Heterogeneitiesmentioning

confidence: 99%

Modelling resource heterogeneities in cloud simulations and quantifying their accuracy

Zakarya

Gillam

2019

Simulation Modelling Practice and Theory

Self Cite

View full text Add to dashboard Cite

“…Their work guarantees that all the tasks fulfil their deadlines with reduced system power consumption. Servers or processing units are the most power consuming equipment's in datacentres [20]. It is very clear from our previous figures that the ratio of power is very small, when a CPU is 100% utilized and when it is idle.…”

Section: Related Workmentioning

confidence: 86%

“…Our future work will investigate these types of scheduling impacts over energy efficiency, workload performance, and users' monetary costs [20], [50].…”

Section: Discussionmentioning

confidence: 99%

An Energy and Performance Aware Scheduler for Real-Time Tasks in Cloud Datacentres

Ali

Qureshi

et al. 2020

IEEE Access

Self Cite

View full text Add to dashboard Cite

Datacentres provide the foundations for cloud computing, but require large amounts of electricity for their operation. Approaches that promise to reduce power use by minimizing execution time, for example using different scheduling and resource management techniques, are discussed in the literature. This paper summarizes some of the most important scheduling techniques in clouds focusing on power consumption, covering VM-level, host-level and task-level scheduling where the most promising approach is task level scheduling, with energy savings by means of load filtering, consolidation, adapted CPU throughput, or host power control. We explore use of the rate monotonic (RM) and backfilling algorithms for real-time task scheduling in cloud environment because RM is the simplest fixed priority scheduling technique, and thus the choice for modern real-time systems, and prior uses of RM in task scheduling have demonstrated power efficiency with optimal results. We specifically consider deadlinebased tasks scheduling for real-time clouds which, to the best of our knowledge, has not been employed previously. RM with backfilling is experimentally evaluated and results show that, compared to the classical algorithms, all tasks were scheduled with minimum power consumption (5.5%-29.3%), on minimum resources (3.9%-25.2% less) while majority were meeting their deadlines (93.21%-94.7%). The approach can guarantee deadline oriented Software as a Service (SaaS) in cloud if arrival rate i.e. network transfer time can be estimated in advance. We subsequently provided an extension of the proposed approach to task-based load balancing for almost balanced resource utilization and approximately 1.0% to 1.6% energy efficiency.

show abstract

“…In the future, we aim to develop cloud-based systems know as Urban Flow that can check the real-time crowd flow prediction. If we implement the training module on the cloud and then the prediction model is implemented on the edge computing or small data centre, then the benefit is that each vehicle should quickly predict and take appropriate decisions for re-routing to avoid crowed or traffic congestion [38], [39], [40], [41], [42], [43]. For that, we need to deploy fog devices on local regions such as shopping malls or mobile base stations, etc.…”

Section: Discussionmentioning

confidence: 99%

Modeling Dynamic Spatio-Temporal Correlations for Urban Traffic Flows Prediction

Awan¹,

Ahmad

Khan

et al. 2021

IEEE Access

Self Cite

View full text Add to dashboard Cite

Prediction of traffic crowd movement is one of the most important component in many applications' domains ranging from urban management to transportation schedule. The key challenge of citywide crowd flows prediction is how to model spatial and dynamic temporal correlation. However, in recent years several studies have been done, but they lack the ability to effectively and simultaneously model spatial and temporal dependencies among traffic crowd flows. To address this issue, in this paper a novel spatio-temporal deep hybrid neural network proposed termed STD-Net to forecast citywide crowd traffic flows. More specifically, STD-Net contains four major branches, i.e., closeness, period volume, weekly volume, and external branches, respectively. We design a residual neural network unit for each property to depict the spatio-temporal features of traffic flows. For various branches, STD-Net provides distinct weights and then combines the outputs of four branches together. Extensive experiments on two large-scale datasets from New York bike and Beijing taxi have demonstrated that STD-Net achieves competitive performances the existing state-of-the-art prediction baselines. INDEX TERMSDeep learning, Urban crowd flows, Neural networks, Long short-term memory, Convolutional neural network

show abstract

Managing energy, performance and cost in large scale heterogeneous datacenters using migrations

Cited by 27 publications

References 43 publications

Modelling resource heterogeneities in cloud simulations and quantifying their accuracy

Modelling resource heterogeneities in cloud simulations and quantifying their accuracy

An Energy and Performance Aware Scheduler for Real-Time Tasks in Cloud Datacentres

Modeling Dynamic Spatio-Temporal Correlations for Urban Traffic Flows Prediction

Contact Info

Product

Resources

About