Multi-Dimensional Regression Host Utilization algorithm (MDRHU) for Host Overload Detection in Cloud Computing

El-Moursy, Ali A.; Abdelsamea, Amany; Kamran, Rukshanda; Saad, Mohamed

doi:10.1186/s13677-019-0130-2

Cited by 25 publications

(11 citation statements)

References 21 publications

(23 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In recent years, various studies have investigated energy efficiency in cloud data centers [6,10,[18][19][20][21][22][23][24][25][26][27][28][29][30]. The majority of the existing VM consolidation methods only focus on balancing the trade-off between the SLA performance desired by the customers and the energy consumed by the servers [6,24,31].…”

Section: Lecture Reviewmentioning

confidence: 99%

See 1 more Smart Citation

DBSHD: Dynamic Burst-aware Source Host Detection in the Cloud

Rahmani

Khajehvand

Torabian

2022

Preprint

View full text Add to dashboard Cite

One of the essential phases during the Virtual Machine (VM) consolidation is to detect the source (underloaded and overloaded) hosts. Bursts with a sudden and transient increase in the workload have an adverse effect on this process. Also, burst-aware methods for detecting the overloaded and underloaded hosts play a key role in achieving performance and energy efficiency tradeoffs in the cloud computing systems. In this study, we deal with these methods as an open issue in the resource management process. We present a dynamic burst-aware method for detecting the overloaded and underloaded hosts. The principal idea in our proposed method is to keep the weighted average load of the hosts within a dynamic range. The thresholds and weights (impact factors) are obtained from the analysis of the historical data related to hosts. Furthermore, we have used the CloudSim simulator and PlanetLab real-world dataset in order to measure the performance of our proposed method. The simulation results verify the superiority of the proposed technique in achieving performance and energy efficiency tradeoffs, compared to the benchmark methods.

show abstract

Section: Lecture Reviewmentioning

confidence: 99%

“…They used a multiple regression method to identify overloaded hosts. El-Moursy et al [22] improved the algorithm presented in [44].…”

Section: Shaw and Singhmentioning

confidence: 99%

DBSHD: Dynamic Burst-aware Source Host Detection in the Cloud

Rahmani

Khajehvand

Torabian

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…It is applicable for predicting the overloaded hosts only. Moursy et al [9] presented a paper on detecting host overload in cloud computing. The CPU utilization is evaluated through Multi-Dimensional Regression Host Utilization (MDRHU) algorithms via Euclidean distance, and it improves 12% of the energy metric in Overloaded.…”

Section: Load Balancing Approachmentioning

confidence: 99%

Novel Load Balancing Approach Towards Cloud Computing by Reducing Sla Degradation and Energy Consumption

Devi¹,

Kumar²

2022

INDJCSE

View full text Add to dashboard Cite

The advent development of cloud computing is an on-demand adaptive technology for many IT organizations due to its dynamic scalability and virtualization resources as a global service. There are numerous energy-conscious approaches in existence that attempt to minimize energy consumption and Service Level Agreements (SLAs) degradation of the host. In this paper, we proposed a novel load balancing approach that identifies overloaded or underloaded hosts, and selects the VMs (Virtual Machines) for allocation to the host based on the predicted load. We consider CPU utilization parameter for evaluating host loading which is far enough to compute all the characteristics of Physical Machines (PM) provided to the Linear Adam Algorithm for Overloaded detection followed by Interquartile Range (IQR) method for underloaded detection. The Power-Aware Best Fit Decreasing (PABFD) algorithm sorts the VM in descending order based on CPU utilization. Then those filtered VM is allocated to the corresponding host thereby reducing the energy consumption and SLAs degradation. We evaluate the proposed algorithm through CloudSim-3.0.3 simulation and random datasets with different workloads on a real Planet Lab. After the evaluation, the simulation result of the proposed method indicates a significant performance by reducing the metric parameters such as SLA degradation, number of VM migration, the total number of host shutdowns, and reduce energy consumption.

show abstract

“…They have developed a discrete differential evolution heuristic to explore the global optimum solution for VM assignment. El-Moursy et al [30] designed a multi-dimension-based regression algorithms to identify overloaded PMs, which integrate memory, CPU, network BW usage. Ranjbari et al [31] suggested a technique of learning automata to improve resource usage while reducing energy consumption.…”

Section: Vm Consolidationmentioning

confidence: 99%

Deep learning-based multivariate resource utilization prediction for hotspots and coldspots mitigation in green cloud data centers

2021

View full text Add to dashboard Cite

Dynamic virtual machine (VM) consolidation is a constructive technique to enhance resource usage and is extensively employed to minimize data centers' energy consumption. However, in the current approaches, consolidation techniques are heavily relied on reducing the actively used physical servers (PMs) based on their current resource utilization without considering future resource demands. Also, many of the reported works for cloud workload prediction applied univariate time series-based forecasting models and neglected the dependency of other resource utilization metrics. Thus, resulting in inaccurate predictions, unnecessary migrations, high migration costs, and increased service level agreement violations (SLAVs) may nullify the consolidation benefits. To efficiently address this issue, we propose a multivariate resource usage prediction-based hotspots and coldspots mitigation approach that considers both the current and future usage of resources with O(sk) time complexity, where s and k denote the number of PMs and VMs, respectively. The proposed technique uses a clustering-based stacked bidirectional (Long Short-Term Memory) LSTM deep learning network to predict the future memory and CPU usage of PMs and VMs with high accuracy and O((Q(Q + W) * Θ) computational complexity, where Q, W, and Θ represent the number of hidden layer cells, outputs, and training epochs, respectively. Through extensive simulations based on Google's cluster workload traces, we demonstrate that our proposed method obtains substantial improvements in terms of prediction performance, energy-efficiency, actively used PMs, VM migrations, and SLA violations over the benchmark approaches.

show abstract

Multi-Dimensional Regression Host Utilization algorithm (MDRHU) for Host Overload Detection in Cloud Computing

Cited by 25 publications

References 21 publications

DBSHD: Dynamic Burst-aware Source Host Detection in the Cloud

DBSHD: Dynamic Burst-aware Source Host Detection in the Cloud

Novel Load Balancing Approach Towards Cloud Computing by Reducing Sla Degradation and Energy Consumption

Deep learning-based multivariate resource utilization prediction for hotspots and coldspots mitigation in green cloud data centers

Contact Info

Product

Resources

About