Prediction Model for Virtual Machine Power Consumption in Cloud Environments

Veni, T.; Bhanu, S. Mary Saira

doi:10.1016/j.procs.2016.05.137

Cited by 6 publications

(24 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Therefrom, the challenge of system power attribution (problem category P8) was tackled to proceed to guests' power models. 3.1.4 P4: Dependency of power model on workload [60,75,[80][81][82][83][87][88][89][90][91]. This category regards the perceived dependency of a VE's power consumption model on the tasks it is processing.…”

Section: P3mentioning

confidence: 99%

“…1 Adaptation to change in the number of co-hosted, concurrent VEs: is widely achieved through time-division multiplexing of event counters [72,75,82,91,98],…”

Section: A17: Model Adaptation Techniquementioning

confidence: 99%

“…Standard Performance Evaluation Corporation (SPEC) [47,52,56,72,75,87,92,106,131,132,142,177,177,185,186] PrincetonApplication Repository for Shared-Memory Computers (PARSEC) [93,94,182] NAS Parallel Benchmarks (NAS NBP) [68,91,96,182] BitCurrent CloudHarmonics BenchCloud [181] CloudSuite [133,135,181] Yahoo! Cloud Serving Benchmark (YCSB) [181] Transaction Processing Performance Council -Web Benchmark (TPC-W) [181] HiBench [181] Iozone [68,96,142] RuBBoS bulletin board [60] Apache [60,131] Tbench [60] Multi-core Processor Architecture and Communication (MPAC) [79] Iometer [87] NBench (BYTEMARK) [74,96] Fhourstone [74] Dhrystoe [74] Ubench [74] Snort [97] TCPreplay [97] Bonnie++ …”

Section: Bibliography Reference Frequencymentioning

confidence: 99%

“…Vector Machine, Neural Networks) [75,89,91] Category D4 regards models of power consumed by mass storage local to the host system [98]. These models attempt to predict power consumption in terms of activity metrics such as total amount of time spent in a known state (in terms of power consumption, e.g.…”

Section: Models Of Host System Power Consumption: D1: Linear Regression Models; D2: Non-linear Regression Models; D3: Machine-learnt Modementioning

confidence: 99%

“…On the other hand, each one of [75,82,91,93,94] is capable of adapting to different workload types without re-training. All predictors here are event counters.…”

Section: Adaptable Modelsmentioning

confidence: 99%

See 4 more Smart Citations

A Categorization of Power Measurement and Power Consumption Models in the Telco Cloud

Depasquale

Salam

Davoli

2021

Preprint

View full text Add to dashboard Cite

This article surveys the literature, over the period 2010-2020, on measurement of power consumption and relevant power models of virtual entities as they apply to the telco cloud. Hardware power meters are incapable of measuring power consumption of individual virtual entities co-hosted on a physical machine. Thus, software power meters are inevitable, yet their development is difficult. Indeed, there is no direct approach to measurement and, therefore, modeling through proxies of power consumption must be used. In this survey, we present trends, fallacies and pitfalls. Notably, we identify limitations of the widely-used linear models and the progression towards Artificial Intelligence / Machine Learning techniques as a means of dealing with the seven major dimensions of variability: workload type; computer virtualization agents; system architecture and resources; concurrent, co-hosted virtualized entities; approaches towards attribution of power consumption to virtual entities; frequency; and temperature.

show abstract

Section: P3mentioning

confidence: 99%

“…1 Adaptation to change in the number of co-hosted, concurrent VEs: is widely achieved through time-division multiplexing of event counters [72,75,82,91,98],…”

Section: A17: Model Adaptation Techniquementioning

confidence: 99%

Section: Bibliography Reference Frequencymentioning

confidence: 99%

Section: Models Of Host System Power Consumption: D1: Linear Regression Models; D2: Non-linear Regression Models; D3: Machine-learnt Modementioning

confidence: 99%

“…On the other hand, each one of [75,82,91,93,94] is capable of adapting to different workload types without re-training. All predictors here are event counters.…”

Section: Adaptable Modelsmentioning

confidence: 99%

See 3 more Smart Citations