Abstract

<div>Increasing power efficiency is one of the most important operational factors for any data centre providers. In this context, one of the most useful approaches is to reduce the number of utilized Physical Machines (PMs) through optimal distribution and re-allocation of Virtual Machines (VMs) without affecting the Quality of Service (QoS). Dynamic VMs provisioning makes use of monitoring tools, historical data, prediction techniques, as well as placement algorithms to improve VMs allocation and migration. Consequently, the efficiency of the data centre energy consumption increases.</div><div>In this thesis, we propose an efficient real-time dynamic provisioning framework to reduce energy in heterogeneous data centres. This framework consists of an efficient workload preprocessing, systematic VMs clustering, a multivariate prediction, and an optimal Virtual Machine Placement (VMP) algorithm. Additionally, it takes into consideration VM and user behaviours along with the existing state of PMs. The proposed framework consists of a pipeline successive subsystems. These subsystems could be used separately or combined to improve accuracy, efficiency, and speed of workload clustering, prediction and provisioning purposes.<br></div><div>The pre-processing and clustering subsystems uses current state and historical workload data to create efficient VMs clusters. Efficient VMs clustering include less consumption resources, faster computing and improved accuracy. A modified multivariate Extreme Learning Machine (ELM)-based predictor is used to forecast the number of VMs in each cluster for the subsequent period. The prediction subsystem takes users’ behaviour into consideration to exclude unpredictable VMs requests.<br></div><div>The placement subsystem is a multi-objective placement algorithm based on a novel Machine Condition Index (MCI). MCI represents a group of weighted components that is inclusive of data centre network, PMs, storage, power system and facilities used in any data centre. In this study it will be used to measure the extent to which PM is deemed suitable for handling the new and/or consolidated VM in large scale heterogeneous data centres. It is an efficient tool for comparing server energy consumption used to augment the efficiency and manageability of data centre resources.</div><div> The proposed framework components separately are tested and evaluated with both synthetic and realistic data traces. Simulation results show that proposed subsystems can achieve efficient results as compared to existing algorithms. <br></div>

Highlights

  • Several energy minimization strategies can be used in data centres, but the most important of them is done by switching off unused Physical Machines (PMs)

  • Simulation results obtained by proposed Machine Condition Index (MCI) and Virtual Machine Placement (VMP)-based multi-objective algorithms in carefully designed experiments validate its effectiveness, taking into considerations the challenges associated with the resolution of the Virtual Machines (VMs) consolidation problem introduced in this thesis

  • This thesis proposes a comprehensive real-time VM consolidation framework that focuses on energy consumption in large-scale heterogeneous data centres

Read more

Summary

Introduction

The proposed real-time framework uses VMP to map VMs to PMs in order to minimize the number of PMs required by the set of VMs. MCI will be used to effectively formulate the problem of VMP from that of multi-objective optimization into that of a single-objective. By using MCI as an objective function to optimize the VMs placement, the proposed VMP are imbued with the following characteristics:. MCI is adopted as an effective tool for comparing the services available by those providers, and can lead to increased efficiency and manageability of resource usage within an enterprise. MCI can be considered as a standard unit for measuring cloud resources within a data centre IT infrastructure. The main advantages of using MCI as a cloud unit are as follows:.

Context and Motivation
Thesis Objectives
Thesis Contributions
Thesis Organization
Background and Literature Review
Workload Prediction Subsystems
Clustering Process
User and VM Behaviour
Prediction Process
Resources State Subsystems
Host Underload Detection
Host Overload Detection
VM Selection Subsystem
Conventional, non ML-based, Techniques
Machine Learning Techniques
VM Placement (VMP) Subsystems
Deterministic Algorithms
Heuristic Algorithms
Approximation Algorithms
Meta-heuristic
VM Migration
Network Effect
Chapter Summary
A Real-Time Energy-Conserving VM-Provisioning Framework for Cloud-data centres
Energy-Aware Framework Components
Feature Selection and Clustering Subsystem
Mapping and Filtering Process
VM Placement Process
Data Monitoring and Thesis Simulated Data
A Systematic Cloud Workload Clustering in Large-Scale Data Centres
Cloud Workload Clustering
Hierarchical Clustering
Partitional Clustering
Density-based Clustering
Model-based Clustering
Grid-based Clustering
Summary
Proposed VMs/Tasks Clustering Subsystem
Feature Selection Stage
Proposed Pre-processing Stage
Clustering Stage
Recommended Clustering Stage
Selected Validation Indices
Proposed Systematic VMs/tasks Clustering
17: Select the number of clusters
Experimental Evaluation and Comparison
Prediction Subsystem
User Behavior-Based Filtering Process
1: Extract all names in a specific period 2
Observation Window Size
Prediction Window Size
Improved ELM Predictor
Experimental Results
Machine Condition Index (MCI)
1: Given a power measurement vs percentage usage for each MCI element 2
Multi-Objective VMP based MCI
Objective Functions
Input Data
Output Data
Constraints
VMP based Multi-Objective Genetic Algorithm
1: Random initial population P OP0 2
MCI as a Cloud Pricing Unit
Real-Time VM Consolidation Framework
Power Consumption Experiment Results
Conclusions
Future Work
Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.