Search

With the rapid development and application of container cloud computing-related technologies, more and more applications are being deployed to container cloud clusters. As an essential feature of container cloud platforms and cloud-native architecture, auto-scaling aims to automatically and quickly adjust the allocation of cloud resources according to the resource requirements of applications. Currently, widely used responsive auto-scaling methods, such as Kubernetes HPA, exhibit certain lags due to the startup time costs of containers and Pods. This lag makes it difficult to guarantee the service quality of applications when there is a sudden increase in online application load. This paper proposes a dynamic interval auto-scaling optimization method based on Informer time series prediction. By predicting online application load and dynamically determining the auto-scaling interval, sufficient resources are allocated to the application in advance. In the experiments conducted on the official World Cup forum load and Alibaba cluster CPU load, the Informer time series prediction algorithm demonstrated better long-sequence time series prediction capabilities compared to algorithms such as LSTM and RNN. In elastic scaling experiments, compared to Kubernetes HPA, the method proposed in this paper reduces the average application response time from 0.821 seconds to 0.692 seconds, and the SLA violation rate decreases from 18.277% to 9.157%. This indicates a significant improvement in the service quality metrics of online applications. Furthermore, the proposed method effectively maintains a balance between high CPU resource utilization and low application response time and SLA violation rate, which is something RNN-based elastic scaling method cannot achieve, as it can only reduce application response time and SLA violation rate by sacrificing CPU resource utilization.

Auto-scaling systems in cloud computing are important for handling application workload fluctuations. This research uses machine learning to predict resource requirements based on workload work patterns and design an automatic scaling system. The dataset used includes features of node name, time, CPU usage percentage, and RAM usage. The ML model is applied for prediction regression of CPU usage percentage, CPU load, and RAM usage, and then server workload is classified into four categories: Very High, High, Low, and Very Low. The autoscaling system used is horizontal scaling. From the results of this research, it was found that the stacking algorithm with the base learner Random Forest and XGBoost had better performance in producing predictive regression. Then, after performing stability testing using K-Fold cross-validation by classifying based on workload status, it was found that the Gradient Boosting algorithm had better results compared to other algorithms, namely for the percentage of CPU usage with an accuracy of 0.998, precision 0.9, recall 0.878, f1score 0.888; CPU load average 15 minutes with accuracy 0.997, precision 0.854, recall 0.863, f1score 0.863; Meanwhile, the percentage of RAM usage is accuracy 0.992, precision 0.986, recall 0.986, and f1score 0.986. However, the XGBoost algorithm also has test results that are almost the same as Gradient Boosting.

In the dynamic world of cloud computing, auto-scaling stands as a beacon of efficiency, dynamically aligning resources with fluctuating demands. This paper presents a comprehensive review of auto-scaling techniques, highlighting significant advancements and persisting challenges in the field. First, we overview the fundamental principles and mechanisms of auto-scaling, including its role in improving cost efficiency, performance, and energy consumption in cloud services. We then discuss various strategies employed in auto-scaling, ranging from threshold-based rules and queuing theory to sophisticated machine learning and time series analysis approaches. After that, we explore the critical issues in auto-scaling practices and review several studies that demonstrate how these challenges can be addressed. We then conclude by offering insights into several promising research directions, emphasizing the development of predictive scaling mechanisms and the integration of advanced machine learning techniques to achieve more effective and efficient auto-scaling solutions. [ABSTRACT FROM AUTHOR]

Traditional auto-scaling approaches are conceived as reactive automations, typically triggered when predefined thresholds are breached by resource consumption metrics. Managing such rules at scale is cumbersome, especially when resources require non-negligible time to be instantiated. This paper introduces an architecture for predictive cloud operations, which enables orchestrators to apply time-series forecasting techniques to estimate the evolution of relevant metrics and take decisions based on the predicted state of the system. In this way, they can anticipate load peaks and trigger appropriate scaling actions in advance, such that new resources are available when needed. The proposed architecture is implemented in OpenStack, extending the monitoring capabilities of Monasca by injecting short-term forecasts of standard metrics. We use our architecture to implement predictive scaling policies leveraging on linear regression, autoregressive integrated moving average, feed-forward, and recurrent neural networks (RNN). Then, we evaluate their performance on a synthetic workload, comparing them to those of a traditional policy. To assess the ability of the different models to generalize to unseen patterns, we also evaluate them on traces from a real content delivery network (CDN) workload. In particular, the RNN model exhibites the best overall performance in terms of prediction error, observed client-side response latency, and forecasting overhead. The implementation of our architecture is open-source.

The synergy of edge computing and Machine Learning (ML) holds immense potential for revolutionizing Internet of Things (IoT) applications, particularly in scenarios characterized by high-speed, continuous data generation. Offline ML algorithms struggle with streaming data as they rely on static datasets for model construction. In contrast, Online Machine Learning (OML) adapts to changing environments by training the model with each new observation in real-time. However, developing OML algorithms introduces complexities such as bias and variance considerations, making the selection of suitable estimators challenging. In this challenging landscape, ensemble learning emerges as a promising approach, offering a strategic framework to navigate the bias-variance tradeoff and enhance prediction accuracy by amalgamating outputs from diverse ML models. This paper introduces a novel ensemble method tailored for edge computing environments, designed to efficiently operate on resource-constrained devices while accommodating various online learning scenarios. The primary objective is to enhance predictive accuracy at the edge, thereby empowering IoT applications with robust decision-making capabilities. Our study addresses the critical challenges of ML in resource-constrained edge computing environments, offering practical insights for enhancing predictive accuracy and scalability in IoT applications. To validate our ensemble's efficacy, we conducted comprehensive experimental evaluations leveraging both synthetic and real-world datasets. The results indicate that our ensemble surpassed state-of-the-art data stream algorithms and ensemble regressors across a range of regression metrics, underlining its superior predictive prowess. Furthermore, we scrutinized the ensemble's performance within the realm of auto-scaling for Virtual Network Function (VNF)-based applications situated at the network's edge, thereby elucidating its applicability and scalability in real-world scenarios. [ABSTRACT FROM AUTHOR]

A recent development in cloud computing has introduced serverless technology, enabling the convenient and flexible management of cloud-native applications. Typically, the Function-as-a-Service (FaaS) solutions rely on serverless backend solutions, such as Kubernetes (K8s) and Knative, to leverage the advantages of resource management for underlying containerized contexts, including auto-scaling and pod scheduling. To take the advantages, recent cloud service providers also deploy self-hosted serverless services by facilitating their on-premise hosted FaaS platforms rather than relying on commercial public cloud offerings. However, the lack of standardized guidelines on K8s abstraction to fairly schedule and allocate resources on auto-scaling configuration options for such on-premise hosting environment in serverless computing poses challenges in meeting the service level objectives (SLOs) of diverse workloads. This study fills this gap by exploring the relationship between auto-scaling behavior and the performance of FaaS workloads depending on scaling-related configurations in K8s. Based on comprehensive measurement studies, we derived the logic as to which workload should be applied and with what type of scaling configurations, such as base metric, threshold to maximize the difference in latency SLO, and number of responses. Additionally, we propose a methodology to assess the scaling efficiency of the related K8s configurations regarding the quality of service (QoS) of FaaS workloads. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Engineering Sciences & Information Technology is the property of Arab Journal of Sciences & Research Publishing (AJSRP) and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Cloud computing has revolutionized the way software, platforms, and infrastructure can be acquired by making them available as on-demand services that can be accessed from anywhere via a web browser. Due to its ubiquitous nature Cloud data centers continuously experience fluctuating workloads which demands for dynamic resource provisioning. These workloads are either placed on Virtual Machines (VMs) or containers which abstract the underlying physical resources deployed at the data center. A proactive or reactive method can be used to allot required resources to the workload. Reactive approaches tend to be inefficient as it takes a significant amount of time to configure the resources to meet the change in demands. A proactive approach for resource management is better in meeting workload demands as it makes an appropriate number of resources available in advance to cater to the fluctuations in workload. The success of such an approach relies on the ability of the resource management module of a data center to accurately predict future workloads. Machine Learning (ML) has already proven itself to be very effective in performing prediction in various domains. In this work, we propose an ML meta-classifier based on stacked generalization for predicting future workloads utilising the past workload trends which are recorded as event logs at Cloud data centers. The proposed model showed a prediction accuracy of 98.5% indicating its applicability for the Cloud environment where SLA requirements must be closely adhered to. [ABSTRACT FROM AUTHOR]

Containerization has become a widely popular virtualization mechanism alongside Virtual Machines (VMs) to deploy applications and services in the cloud. Containers form the backbone of the modern architectures around microservices and provide a lightweight virtualization mechanism for IoT and Edge systems. Elasticity is one of the key requirements of modern applications with various constraints ranging from Service Level Agreements (SLA) to optimization of resource utilization, cost management, etc. Auto Scaling is a technique used to attain elasticity by scaling the number of containers or resources. This work introduces a novel mechanism for auto-scaling containers in cloud environments, addressing the key elasticity requirement in modern applications. The proposed mechanism combines predictive analysis using the Auto-Regressive Integrated Moving Average (ARIMA) model and control theory utilizing the Proportional-Integral-Derivative (PID) controller. The major contributions of this work include the development of the ARIMA-PID algorithm for forecasting resource utilization and maintaining desired levels, comparing ARIMA-PID with existing threshold mechanisms, and demonstrating its superior performance in terms of CPU utilization and average response times. Experimental results showcase improvements of approximately 10% in CPU utilization and 30% [ABSTRACT FROM AUTHOR]

Summary: The expeditious deployment of Cloud applications and services on wide‐ranging Cloud Data Centres (CDC) gives rise to the utilization of many resources. Moreover, by the increase in resource utilization, virtualization also greatly impacts achieving desired performance. The major challenges in virtualization are detecting over‐utilized or under‐utilized hosts at the right time and the proper scaling of Virtual Machines (VM) on the accurate host. Auto‐scaling in Cloud Computing allows the service providers to scale up or down the resources automatically and provides on‐demand computing power and storage capacities. Effective utilization and autonomous scaling of resources eventually reduce the load, energy consumption, and operating costs. In this paper, an efficient auto‐scaling approach for predicting host load through VM migration has been proposed. The ensemble method using different time‐series forecasting models has been proposed to forecast the approaching workload on the host. Based on this predicted load, different algorithms have been devised to detect over‐utilized and under‐utilized hosts and VMs can be migrated. The designed approach has been validated by experimentation on a real‐time Google cluster dataset. The proposed technique significantly improves average CPU utilization and reduces over‐utilization and under‐utilization. It also minimizes response time, service level agreement violations, and the slighter number of migrations and scaling overhead. [ABSTRACT FROM AUTHOR]

In recent years, the demand for real-time data processing has been increasing, and various stream processing systems have emerged. When the amount of data input to the stream processing system fluctuates, the computing resources required by the stream processing job will also change. The resources used by stream processing jobs need to be adjusted according to load changes, avoiding the waste of computing resources. At present, existing works adjust stream processing jobs based on the assumption that there is a linear relationship between the operator parallelism and operator resource consumption (e.g., throughput), which makes a significant deviation when the operator parallelism increases. This paper proposes a nonlinear model to represent operator performance. We divide the operator performance into three stages, the Non-competition stage, the Non-full competition stage, and the Full competition stage. Using our proposed performance model, given the parallelism of the operator, we can accurately predict the CPU utilization and operator throughput. Evaluated with actual experiments, the prediction error of our model is below 5%. We also propose a quick accurate auto-scaling (QAAS) method that uses the operator performance model to implement the auto-scaling of the operator parallelism of the Flink job. Compared to previous work, QAAS is able to maintain stable job performance under load changes, minimizing the number of job adjustments and reducing data backlogs by 50%. [ABSTRACT FROM AUTHOR]

Data Stream Processing (DSP) applications analyze data flows in near real-time by means of operators, which process and transform incoming data. Operators handle high data rates running parallel replicas across multiple processors and hosts. To guarantee consistent performance without wasting resources in the face of variable workloads, auto-scaling techniques have been studied to adapt operator parallelism at run-time. However, most of the effort has been spent under the assumption of homogeneous computing infrastructures, neglecting the complexity of modern environments. We consider the problem of deciding both how many operator replicas should be executed and which types of computing nodes should be acquired. We devise heterogeneity-aware policies by means of a twolayered hierarchy of controllers. While application-level components steer the adaptation process for whole applications, aiming to guarantee user-specified requirements, lower-layer components control auto-scaling of single operators. We tackle the fundamental challenge of performance andworkload uncertainty, exploiting Bayesian optimization (BO) and reinforcement learning (RL) to devise policies. The evaluation shows that our approach is able to meet users' requirements in terms of response time and adaptation overhead, while minimizing the cost due to resource usage, outperforming state-of-the-art baselines. We also demonstrate how partial model information is exploited to reduce training time for learning-based controllers. [ABSTRACT FROM AUTHOR]

Recently, microservices have become a commonly-used architectural pattern for building cloud-native applications. Cloud computing provides flexibility for service providers, allowing them to remove or add resources depending on the workload of their web applications. If the resources allocated to the service are not aligned with its requirements, instances of failure or delayed response will increase, resulting in customer dissatisfaction. This problem has become a significant challenge in microservices-based applications, because thousands of microservices in the system may have complex interactions. Auto-scaling is a feature of cloud computing that enables resource scalability on demand, thus allowing service providers to deliver resources to their applications without human intervention under a dynamic workload to minimize resource cost and latency while maintaining the quality of service requirements. In this research, we aimed to establish a computational model for analyzing the workload of all microservices. To this end, the overall workload entering the system was considered, and the relationships and function calls between microservices were taken into account, because in a large-scale application with thousands of microservices, accurately monitoring all microservices and gathering precise performance metrics are usually difficult. Then, we developed a multi-criteria decision-making method to select the candidate microservices for scaling. We have tested the proposed approach with three datasets. The results of the conducted experiments show that the detection of input load toward microservices is performed with an average accuracy of about 99% which is a notable result. Furthermore, the proposed approach has demonstrated a substantial enhancement in resource utilization, achieving an average improvement of 40.74%, 20.28%, and 28.85% across three distinct datasets in comparison to existing methods. This is achieved by a notable reduction in the number of scaling operations, reducing the count by 54.40%, 55.52%, and 69.82%, respectively. Consequently, this optimization translates into a decrease in required resources, leading to cost reductions of 1.64%, 1.89%, and 1.67% respectively. [ABSTRACT FROM AUTHOR]

Edge computing has gained widespread adoption for time-sensitive applications by offloading a portion of IoT system workloads from the cloud to edge nodes. However, the limited resources of IoT edge devices hinder service deployment, making auto-scaling crucial for improving resource utilization in response to dynamic workloads. Recent solutions aim to make auto-scaling proactive by predicting future workloads and overcoming the limitations of reactive approaches. These proactive solutions often rely on time-series data analysis and machine learning techniques, especially Long Short-Term Memory (LSTM), thanks to its accuracy and prediction speed. However, existing auto-scaling solutions often suffer from oscillation issues, even when using a cooling-down strategy. Consequently, the efficiency of proactive auto-scaling depends on the prediction model accuracy and the degree of oscillation in the scaling actions.This paper proposes a novel approach to improve prediction accuracy and deal with oscillation issues. Our approach involves an automatic featurization phase that extracts features from time-series workload data, improving the prediction’s accuracy. These extracted features also serve as a grid for controlling oscillation in generated scaling actions. Our experimental results demonstrate the effectiveness of our approach in improving prediction accuracy, mitigating oscillation phenomena, and enhancing the overall auto-scaling performance.

INTRODUCTION: Cloud services are becoming increasingly important as advanced technology changes. In these kinds of cases, the volume of work on the corresponding server in public real-time data virtualized environment can vary based on the user's needs. Cloud computing is the most recent technology that provides on-demand access to computer resources without the user's direct interference. Consequently, cloud-based businesses must be scalable to succeed. OBJECTIVES: The purpose of this research work is to describe a new virtual cluster architecture that allows cloud applications to scale dynamically within the virtualization of cloud computing scale using auto-scaling, resources can be dynamically adjusted to meet multiple demands of the customers. METHODS: An auto-scaling algorithm based on the current implementation sessions will be initiated for automated provisioning and balancing of virtualized resources. The suggested methodology also considers the cost of energy. RESULTS: The proposed research work has shown that the suggested technique can handle sudden load demands while maintaining higher resource usage and lowering energy costs efficiently. CONCLUSION: Auto-scaling features are available in measures in order groups, allowing you to automatically add or remove instances from a managed instance group based on changes in load. This research work provides an analysis of autoscaling mechanisms in cloud services that can be used to find the most efficient and optimal solution in practice and to manage cloud services efficiently. [ABSTRACT FROM AUTHOR]

The resource usage behaviors of application workloads are currently the primary concern of cloud providers offering hosting services. These services should be able to adapt to workload changes by automatically provisioning and de-provisioning resources so that, at all times, the existing resources in a system match the current service demand. Such behavior can be achieved manually by hiring a DevOps team to manage the application’s resources. Another option would be automating the resource provisioning processing using automated rules. Once such rules are met, the hosting environment will scale the resources accordingly. However, managing a DevOps team or creating flaky rules can lead to over-scaling application resources. This work proposes a new approach: a proactive auto-scaling framework built on an ensemble model. Such a model utilizes several machine learning techniques to scale application resources to match resource demand before the need arises. We evaluated our solution against three real production applications hosted on Cegedim Cloud Hosting Environment, an industrial environment serving several cloud applications from various domains, and against other machine learning models used in similar proactive auto-scaling experiments mentioned in past work. The experimentation results show that predicting application resources like CPU or RAM is feasible. Moreover, even in production environments, our ensemble model performs optimally in the CPU case and is near the optimal model when predicting RAM resources.

On-demand availability and resource elasticity features of Cloud computing have attracted the focus of various research domains. Mobile cloud computing is one of these domains where complex computation tasks are offloaded to the cloud resources to augment mobile devices’ cognitive capacity. However, the flexible provisioning of cloud resources is hindered by uncertain offloading workloads and significant setup time of cloud virtual machines (VMs). Furthermore, any delays at the cloud end would further aggravate the miseries of real-time tasks. To resolve these issues, this paper proposes an auto-scaling framework (ACF) that strives to maintain the quality of service (QoS) for the end users as per the service level agreement (SLA) negotiated assurance level for service availability. In addition, it also provides an innovative solution for dealing with the VM startup overheads without truncating the running tasks. Unlike the waiting cost and service cost tradeoffbased systems or threshold-rule-based systems, it does not require strict tuning in the waiting costs or in the threshold rules for enhancing the QoS. We explored the design space of the ACF system with the CloudSim simulator. The extensive sets of experiments demonstrate the effectiveness of the ACF system in terms of good reduction in energy dissipation at the mobile devices and improvement in the QoS. At the same time, the proposed ACF system also reduces the monetary costs of the service providers. [ABSTRACT FROM AUTHOR]

The dynamic nature of the cloud environment increases the complexity of managing its resources and the distribution of user workload between the available containers in the data center. However, the workload must be balanced to improve the cloud system's overall performance. Generally, most of the existing load balancing techniques suffer from performance degradation due to the communication overheads among the containers. Moreover, less attention is given to stabilize the load in a multi-cloud environment. Therefore, to overcome this problem, there is a need to develop an elastic load balancing method to improve the performance of cloud systems. This paper proposed an autonomic CSO-ILB load balancer to ensure the elasticity of the cloud system and balance the user workload among the available containers in a multi-cloud environment. The concept of multi-loop has been utilized in our approach to enabling efficient self-management before load balancing. The tasks are scheduled to the containers using an extended scheduling algorithm called Deadline-Constrained Make-span Minimization for Multi-Task Scheduling (DCMM-MTS). Based on the task scheduling, the load in each container is computed and then balanced using the proposed load balancer algorithm CSO-ILB. The proposed approach is evaluated in the Container CloudSim platform, and the performance is compared with the existing meta-heuristic algorithms such as Ant Colony Optimization, Bee Colony Optimization, Shuffled Frog Leaping Algorithm and Cat Swarm Optimization (CSO). The simulations proved that the proposed approach outperformed the other approaches in terms of reliability, CPU utilization, make-span, energy utilization, response time, execution cost, idle time, and task migration. [ABSTRACT FROM AUTHOR]

Cloud providers tend to support dynamic computing resources reallocation (e.g., Autoscaling) to handle the bursty workload for web applications (e.g., e-commerce) in the cloud environment. Nevertheless, we demonstrate that directly scaling a bottleneck server without quickly adjusting its soft resources (e.g., server threads and database connections) can cause significant response time fluctuations of the target web application. Since soft resources determine the request processing concurrency of each server in the system, simply scaling out/in the bottleneck service can unintentionally change the concurrency level of related services, inducing either under- or over-utilization of the critical hardware resource. In this paper, we propose the Scatter-Concurrency-Throughput (SCT) model, which can rapidly identify the near-optimal soft resource allocation of each server in the system using the measurement of each server’s real-time throughput and concurrency. Furthermore, we implement a Concurrency-aware autoScaling (ConScale) framework that integrates the SCT model to quickly reallocate the soft resources of the key servers in the system to best utilize the new hardware resource capacity after the system scaling. Based on extensive experimental comparisons with two widely used hardware-only scaling mechanisms for web applications: EC2-AutoScaling (VM-based autoscaler) and Kubernetes HPA (container-based autoscaler), we show that ConScale can successfully mitigate the response time fluctuations over the system scaling phase in both VM-based and container-based environments. [ABSTRACT FROM AUTHOR]

Microservices are being used by businesses to split monolithic software into a set of small services whose instances run independently in containers. Load balancing and auto-scaling are important to cloud features for cloud-based container microservices because they control the number of resources available. The current issues concerning load balancing and auto-scaling techniques in Cloud-based container microservices were investigated in this paper. Server overloaded, service failure and traffic spikes were the key challenges faced during the microservices communication phase, making it difficult to provide better Quality of Service (QoS) to users. The aim is to critically investigate the addressed issues related to Load balancing and Auto-scaling in Cloudbased Container Microservices (CBCM) in order to enhance performance for better QoS to the users. [ABSTRACT FROM AUTHOR]

Cloud computing has been consolidated as a support for the vast majority of current and emerging technologies. However, there are some barriers that prevent the exploitation of the full potential of this technology. First, the major cloud providers currently put the onus of implementing the mechanisms that ensure compliance with the desired service levels on cloud consumers. However, consumers do not have the required expertise. Since each cloud provider exports a different set of low‐level metrics, the strategies defined to ensure compliance with the established service‐level agreement (SLA) are bound to a particular cloud provider. This fosters provider lock‐in and prevents consumers from benefiting from the advantages of multi‐cloud environments. This paper presents a solution to the problem of automatically translating SLAs into objectives expressed as metrics that can be measured across multiple cloud providers. First, we propose an intelligent knowledge‐based system capable of automatically translating high‐level SLAs defined by cloud consumers into a set of conditions expressed as vendor‐neutral metrics, providing feedback to cloud consumers (intelligent tutoring system). Secondly, we present the set of vendor‐neutral metrics and explain how they can be measured for the different cloud providers. Finally, we report a validation based on two use cases (IaaS and PaaS) in a multi‐cloud environment formed by leading cloud providers. This evaluation has demonstrated that, thanks to the complementarity of the two solutions, cloud consumers can automatically and transparently exploit the multi‐cloud in many application domains, as endorsed by the cloud experts consulted in the course of this study. [ABSTRACT FROM AUTHOR]

Scale the ionosonde ionograms to produce accurate readings is a professional manual scaling technique. However, there is a high demand for auto-scaling software that can manage a large number of ionograms in order to avoid the time and effort involved in manual scaling as well as human errors. Noise-free, accurate trace identification and precise segmentation are required for the auto-scaling program to work. The Canadian Advanced Digital Ionosonde (CADI) ionograms are processed and auto-scaled using a new model on an open-source (Python) platform in this paper. Filtering the noise, Convolution Neural Network (CNN) based trace detection, layer-wise segmentation, and then extracting the ionospheric features are used to accomplish the scaling accuracy. The investigation uses raw ionogram files generated by the CADI system in Hyderabad, India (Lat: $17.47^{\circ }\text{N}$ , Long: $78.57^{\circ }\text{E}$ ) between 2014 and 2015. Raw ionograms in $^\ast $ .md4 or $^\ast $ .md2 file formats can be accepted by the suggested model (Individual or Hourly integrated). The proposed auto-scaling software tool’s individual block performance is examined with several classes of ionograms, and the overall performance is evaluated with a huge set of ionograms obtained during adverse space weather circumstances (16th to 18th March 2015). Univap Digital Ionosonde Data Analysis (UDIDA) software tool was considered for manual scaling. The results of manual scaling are compared with that of proposed scaling software. In fmin and h’f, respectively, the proposed model has a mean absolute error (MAE) of 0.36 MHz and 11.72 km, and a root mean square error (RMSE) of 0.7 MHz and 22.36 km.

With increasing growth in IoT, the number of devices connected to the Internet is constantly growing. Moreover, the increase in the volume of data and their transmission through the Internet of Things, as well as the existence of inadequate bandwidth, limits cloud-based storage and data processing. Both fog and cloud computing provide the storage space, application, and data for users; however, fog is more proximate to the end user with wider geographical distribution. When bringing the computing resources closer to the required location in the fog environment, the efficiency of the system increases, and the distance at which data must be transmitted decreases. On the other hand, implementing IoT applications and satisfying the requests of end users in fog computing will create new challenges in resource allocation and dynamic resource provisioning. The flexible and usually automatic mechanisms require the determination of required virtual resources to minimize the resource consumption and service level agreement (SLA). In this paper, we introduce a framework for increasing resource management efficiency in the IoT ecosystem based on deep reinforcement learning (DRL). The proposed deep neural network (DNN) method for estimating value functions improves adaptability to different oscillating conditions, learns past sensible strategies, and as a self-learning adaptive system by replicating interactions with the fog environment. The DRL algorithm finds the best destination for implementing IoT services to compromise between minimizing average power consumption, minimizing average service latency, reducing costs, and balancing resource allocation. Finally, through simulations, we show that under different loading rates, the policy used compared to other comparable solutions is to increase utilization and reduce the rate of delay, while ensuring an acceptable level of service quality. [ABSTRACT FROM AUTHOR]

Network Function Virtualization (NFV) offers an alternate method to design, deploy and manage network services. The NFV decouples network functions from the dedicated hardware and moves them to the virtual servers so that they can run in the software. One of the major strengths of the NFV is its ability to dynamically extend or reduce resources allocated to Virtual Network Functions (VNF) as needed and at run-time. There is a need for a comprehensive metering component in the cloud to store and process the metrics/samples for efficient auto-scaling or load-management of the VNF. In this paper, we propose an integrating framework for efficient auto-scaling of VNF using Gnocchi; a time-series database that is integrated within the framework to store, handle and index the time-series data. The objective of this study is to validate the efficacy of employing Gnocchi for auto-scaling of VNF, in terms of aggregated data points, database size, data recovery speed, and memory consumption. The employed methodology is to perform a detailed empirical analysis of the proposed framework by deploying a fully functional cloud to implement NFV architecture using several OpenStack components including Gnocchi. Our results show a significant improvement over the legacy Ceilometer configuration in terms of lower metering storage size, less memory utilization in processing and management of metrics, and reduced time delay in retrieving the monitoring data to evaluate alarms for the auto-scaling of VNF. [ABSTRACT FROM AUTHOR]

Cloud computing has been regarded as a successful paradigm for IT industry by providing benefits for both service providers and customers. In spite of the advantages, cloud computing also suffers from distinct challenges, and one of them is the inefficient resource provisioning for dynamic workloads. Accurate workload predictions for cloud computing can support efficient resource provisioning and avoid resource wastage. However, due to the high-dimensional and high-variable features of cloud workloads, it is difficult to predict the workloads effectively and accurately. The current dominant work for cloud workload prediction is based on regression approaches or recurrent neural networks, which fail to capture the long-term variance of workloads. To address the challenges and overcome the limitations of existing works, we proposed an efficient supervised learning-based Deep Neural Network (esDNN) approach for cloud workload prediction. First, we utilize a sliding window to convert the multivariate data into a supervised learning time series that allows deep learning for processing. Then, we apply a revised Gated Recurrent Unit (GRU) to achieve accurate prediction. To show the effectiveness of esDNN, we also conduct comprehensive experiments based on realistic traces derived fromAlibaba and Google cloud data centers. The experimental results demonstrate that esDNN can accurately and efficiently predict cloud workloads. Compared with the state-of-the-art baselines, esDNN can reduce the mean square errors significantly, e.g., 15%. rather than the approach using GRU only.We also apply esDNN for machines auto-scaling, which illustrates that esDNN can reduce the number of active hosts efficiently, thus the costs of service providers can be optimized. [ABSTRACT FROM AUTHOR]

Internet of Things (IoT) enables smart campuses more convenient for cloud services. The availability of cloud resources to its users appears as a fundamental challenge. The existing research presents several auto-scaling techniques to scale the resources with the increase in users’ demands. However, still, the cloud users of auto-scaled servers experience service disruption, delayed responses, and the occurrence of service bursts. The prevailing burst management framework exhibits limitations in the context of burdening the existing auto-scaled machines for cost estimation and resource allocation. This research presents a 3-axis auto-scaling framework for load balancing and resource allocation by incorporating a dedicated cost estimator and allocator (on the z-axis). The cost estimation server develops a log of existing load estimates of vertical and horizontal servers and scales the new users’ requests in case the vertical threshold is breached with new requests. The cost estimator, in its data structure, keeps track of the current resources available at both vertical and horizontal servers. The historical information of available resources and the new resources’ requests is decided by the cost estimator as per demand and supply scenario. The general characteristics of servers are resources pooling, requests queue development, burst identification, automatic scaling, and load balancing. The cost estimator also prioritizes vertical servers for resource allocations, and switches to the horizontal server when the vertical server reaches its 75% quota of resources. The study simulates 1000 users’ requests of smart campus, adopts state-of-the-art ensemble with bagging strategy and handles an effective class imbalance situation.

The internet of things is an emerging technology used in cloud computing and provides many services of the cloud. The cloud services users mostly suffer from service delays and disruptions due to service cloud resource management based on vertical and horizontal scalable systems. Adding more resources to a single cloud server is called vertical scaling, and an increasing number of servers is known as horizontal scaling. The service-bursts significantly impact the vertical scaled environment where the scale-up degrades the service quality and users’ trust after reaching the server’s maximum capacity. Besides, the horizontally scaled environment, though being resilient, is cost-inefficient. It is also hard to detect and manage bursts online to sustain application efficiency for complex workloads. Burst detection in real-time workloads is a complicated issue because even in the presence of auto-scaling methods, it can dramatically degrade the application’s efficiency. This research study presents a new bursts-aware auto-scaling approach that detects bursts in dynamic workloads using resource estimation, decision-making scaling, and workload forecasting while reducing response time. This study proposes a hybrid auto-scaled service cloud model that ensures the best approximation of vertical and horizontal scalable systems to ensure Quality of Service (QoS) for smart campus-based applications. This study carries out the workload prediction and auto-scaling employing an ensemble algorithm. The model pre-scales the scalable vertical system by leveraging the service-load predictive modeling using an ensemble classification of defined workload estimation. The prediction of the upcoming workload helped scale-up the system, and auto-scaling dynamically scaled the assigned resources to many users’ service requests. The proposed model efficiently managed service-bursts by addressing load balancing challenges through horizontal auto-scaling to ensure application consistency and service availability. The study simulated the smart campus environment model to monitor the time-stamped diverse service-requests appearing with different workloads.