Cloud Workloads Research Articles

A Multiple Controlled Toffoli Driven Adaptive Quantum Neural Network Model for Dynamic Workload Prediction in Cloud Environments.

The key challenges in cloud computing encompass dynamic resource scaling, load balancing, and power consumption. Accurate workload prediction is identified as a crucial strategy to address these challenges. Despite numerous methods proposed to tackle this issue, existing approaches fall short of capturing the high-variance nature of volatile and dynamic cloud workloads. Consequently, this paper introduces a novel model aimed at addressing this limitation. This paper presents a novel Multiple Controlled Toffoli-driven Adaptive Quantum Neural Network (MCT-AQNN) model to establish an empirical solution to complex, elastic as well as challenging workload prediction problems by optimizing the exploration, adaption, and exploitation proficiencies through quantum learning. The computational adaptability of quantum computing is ingrained with machine learning algorithms to derive more precise correlations from dynamic and complex workloads. The furnished input data point and hatched neural weights are refitted in the form of qubits while the controlling effects of Multiple Controlled Toffoli (MCT) gates are operated at the hidden and output layers of Quantum Neural Network (QNN) for enhancing learning capabilities. Complimentarily, a Uniformly Adaptive Quantum Machine Learning (UAQL) algorithm has evolved to functionally and effectually train the QNN. The extensive experiments are conducted and the comparisons are performed with state-of-the-art methods using four real-world benchmark datasets. Experimental results evince that MCT-AQNN has up to 32%-96% higher accuracy than the existing approaches.

Automated Virtual Machine Migration with Load Shifting Technique Avoiding Latency in Cloud Environment

Efficient management of Virtual Machines (VMs) in cloud computing is one key aspect to achieve the best utilization of resources and performance. The study is about improving Automated Virtual Machine Migration with a Load Shifting Technique in order to reduce latency through the cloud domain. This methodology dynamically distribute the load across multiple VMs by migrating their workloads to other physical hosts determined according certain resource utilization metrics available in real-time. The dynamic and unpredictable nature of cloud workloads requires a solution that is more responsive in the way it manages VMs. Live migration and pre-copy migration offer suboptimal performance in terms of time to migrate, downtime due to the prolonged copying period as heaps swell/collapse during live migrations. Utilizing sophisticated algorithms and real-time monitoring to avoid these pitfalls, our methodology allows us for proper load balancing with minimal latency. The primary elements of the proposed framework are a detailed monitoring module, an intelligent load shifting algorithm, high-performance migration method and feedback loop for continuous optimization. The monitoring module is responsible for collecting on-the-fly information about CPU, memory, network traffic and disk I/O such that can be exploited by the load shifting algorithm. The algorithm processes this data to decide which VMs should be migrated with the purpose of ensuring load balancing and decreasing latency achieved using these migrations. In order to test the proposed approach, comprehensive simulations and comparatives analyses utilizing other models were performed. This results in better latency reduction, increased resource utilization and improved overall system performance. In fact, the proposed method can handle not only current challenges in cloud computing such as load balancing and latency issued but also a scalable and effective approach for general cloud infrastructure usage. Finally, the Automated Virtual Machine Migration with Load Shifting Technique proves to be a reliable and successful solution for assisting in managing dynamic cloud workloads. Through real-time monitoring, advanced algorithms and continuous optimization this method has now been able to improve cloud services performance and reliability thereby allowing it to better serve service level agreements as well meet up with user satisfaction. We hope to adapt the algorithm and generalize it for even more advanced predictive models in future iterations, as well as evaluation of practical implementation on cloud-based deployment environments.

Benchmarking and revisiting time series forecasting methods in cloud workload prediction

Benchmarking and revisiting time series forecasting methods in cloud workload prediction

Improving QoS in cloud resources scheduling using dynamic clustering algorithm and SM‐CDC scheduling model

SummaryQuality of Service (QoS) regulates and controls network resources by setting priorities for specific data types. Many clustering algorithms are used to cluster cloud workloads, most of which are static. However, the lack of dynamic algorithms is seen in the face of huge databases that are real‐time and according to the existing clustering conditions. Additionally, fair allocation of tasks on servers and efficient resource utilization pose challenges. In this research, two solutions are proposed to improve the quality of service: the first solution uses the chameleon dynamic algorithm, a method to improve service quality. The chameleon algorithm has been able to show significant performance due to its high accuracy in detecting the smallest distance between clusters. This dynamic algorithm outperforms static algorithms with classification accuracy and response speed, which are the most important parameters of service quality. The second part of the proposed solution is to use the Scheduling Model using Cloud Data Centers (SM‐CDC) system to select the best service provider based on the clustering done in the previous step. A SM‐CDC technique is developed to handle cloud storage center tasks that are stored in electronic devices. According to the comparison with existing scheduling policies, SM‐CDC offered 36% decrease on response time, 50% reduction on cost of resources, and 40% improvement on QoS Satisfaction.

Efficient virtual machine placement in cloud computing environment using BSO-ANN based hybrid technique

Cloud computing has revolutionized the way businesses and individuals access and utilize computing resources. Efficient virtual machine placement is a critical aspect of optimizing resource utilization, reducing operational costs, energy consumption, service level agreement and minimum virtual machine migrations, execution time, and ensuring the overall performance of cloud services. This manuscript introduces a novel approach that combines the creative problem-solving capabilities of brainstorming with the computational power of Artificial Neural Networks (ANN) to address the virtual machine placement problem in cloud environments. In this study, we propose a hybrid technique that leverages the collective intelligence of human brainstorming to generate a diverse set of placement strategies. These strategies are then evaluated, refined, and optimized using an ANN model trained on historical cloud resource allocation workload logs. By integrating the human creative process with the data-driven predictive capabilities of ANN, our approach aims to overcome the limitations of traditional virtual machine placement algorithms, which often struggle to adapt to dynamic workloads and changing resource requirements. The manuscript provides a detailed description of the hybrid technique, including the process of brainstorming for generating placement strategies, data collection and preprocessing, ANN model development, and the integration of these components into an efficient placement system. We present experimental results demonstrating the effectiveness of our approach in optimizing resource allocation, improving service performance, and optimizing resource utilization, reducing energy consumption, service level agreement and minimum virtual machine migrations, reducing execution time compared to existing static, and meta-heuristic methods. The proposed Brain Storming with ANN based Hybrid Technique offers a promising solution for enhancing the efficiency of virtual machine placement in cloud computing environments. BSO-ANN outperforms the existing techniques using performance metrics (energy consumption(Kwh), execution time(ms), SLA violations, and number of migrations). It combines human ingenuity with data-driven insights to adapt to the ever-changing dynamics of cloud workloads. This manuscript contributes to the ongoing research in cloud resource management, offering a practical approach for cloud service providers and organizations to better utilize their resources and enhance the overall quality of cloud-based services. © 2012 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of Global Science and Technology Forum Pte Ltd.

An Advantage Actor-Critic and Proximal Policy Optimization Model for Adaptive Scheduling of Heterogeneous Cloud Workloads

An Advantage Actor-Critic and Proximal Policy Optimization Model for Adaptive Scheduling of Heterogeneous Cloud Workloads

Hybrid deep learning and evolutionary algorithms for accurate cloud workload prediction

Hybrid deep learning and evolutionary algorithms for accurate cloud workload prediction

Multi-objective cuckoo optimizer for task scheduling to balance workload in cloud computing

Multi-objective cuckoo optimizer for task scheduling to balance workload in cloud computing

From SSDs Back to HDDs: Optimizing VDO to Support Inline Deduplication and Compression for HDDs as Primary Storage Media

Deduplication and compression are powerful techniques to reduce the ratio between the quantity of logical data stored and the physical amount of consumed storage. Deduplication can impose significant performance overheads, as duplicate detection for large systems induces random accesses to the backend storage. These random accesses have led to the concern that deduplication for primary storage and HDDs are not compatible. Most inline data reduction solutions are therefore optimized for SSDs and discourage their use for HDDs, even for sequential workloads. In this work, we show that these concerns are valid if and only if the lessons learned from deduplication research are not applied. We have therefore investigated data reduction solutions for primary storage based on the RedHat Virtual Disk Optimizer (VDO) and show that directly applying them can decrease sequential write performance for HDDs by 36×. We then show that slight modifications to VDO plus the integration of a very small SSD area significantly improve performance even beyond the performance without data reduction enabled, making HDDs more cost-efficient for a wide range of mostly sequential cloud workloads than SSDs. Additionally, these VDO optimizations do not require to maintain different code bases for HDDs and SSDs, and we therefore provide the first data reduction solution applicable to both storage media.

OptScaler: A Collaborative Framework for Robust Autoscaling in the Cloud

Autoscaling is a critical mechanism in cloud computing, enabling the autonomous adjustment of computing resources in response to dynamic workloads. This is particularly valuable for co-located, long-running applications with diverse workload patterns. The primary objective of autoscaling is to regulate resource utilization at a desired level, effectively balancing the need for resource optimization with the fulfillment of Service Level Objectives (SLOs). Many existing proactive autoscaling frameworks may encounter prediction deviations arising from the frequent fluctuations of cloud workloads. Reactive frameworks, on the other hand, rely on realtime system feedback, but their hysteretic nature could lead to violations of stringent SLOs. Hybrid frameworks, while prevalent, often feature independently functioning proactive and reactive modules, potentially leading to incompatibility and undermining the overall decision-making efficacy. In addressing these challenges, we propose OptScaler, a collaborative autoscaling framework that integrates proactive and reactive modules through an optimization module. The proactive module delivers reliable future workload predictions to the optimization module, while the reactive module offers a self-tuning estimator for real-time updates. By embedding a Model Predictive Control (MPC) mechanism and chance constraints into the optimization module, we further enhance its robustness. Numerical results have demonstrated the superiority of our workload prediction model and the collaborative framework, leading to over a 36% reduction in SLO violations compared to prevalent reactive, proactive, or hybrid autoscalers. Notably, OptScaler has been successfully deployed at Alipay, providing autoscaling support for the world-leading payment platform.

Towards Carbon Footprint Management in Hybrid Multicloud

Enterprises today aspire to optimize the operating costs and carbon footprint (CFP) of their IT operations jointly without compromising their business imperatives. This has given rise to a hybrid approach in which enterprises retain the dynamic choice to leverage private data centers and one or more public clouds in conjunction. While cloud service providers (CSPs) have long provided APIs for estimating, reconciling, and optimizing operating costs, they have only recently started exposing APIs related to CFP. Indeed, this is a step in the right direction. Nevertheless, our analyses of these APIs reveals many gaps that need to be addressed to facilitate sizing and placement decisions that can factor in carbon. First, there is a lack of standardized, transparent methodology for CFP quantification across different CSPs. Second, the coarse granularity of the CFP data provided today can help with post-facto reporting but is not suitable for proactive fine-grained optimization. Last, enterprises themselves are unable to independently compute the current CFP or estimate potential CFP savings since CSPs do not share the required power usage data. To address these gaps, enterprises have started developing their own carbon assessment methodologies and tools to estimate the CFP of workloads running on public clouds using the available user-facing APIs. These systems hold the promise for an independent and unbiased evaluation and estimation of relative savings between different deployment options by cloud users. We describe and analyze the details of CSP-native carbon-reporting tools and their quantification methodology, and the "outside-of-the-cloud" estimation approaches. Finally, we present opportunities for future research in the direction of trustworthy, fine-grained, public cloud workload CFP estimation, which is a prerequisite for meaningful realization of carbon optimization.

Sustainable energy efficient workflow classification and scheduling in geo distributed cloud datacenter

Data centers are a major source of carbon emissions and are subsequently contributing to global carbon footprints. Keeping in view of providing a sustainable solution to society, we have analyzed various factors that can help to achieve carbon neutrality and maximum sustainability. Our study pointed towards the need to follow a sustainable approach for incoming workflow throughout the life-cycle of Data centers. We analyzed that workloads need to be segregated before assigning them to the data centers so that energy-efficient resource allocation could be done. This paper demonstrates unsupervised learning techniques to cluster the incoming cloud workloads. The heterogeneous workloads were characterized using machine learning approaches and appropriate clusters were crafted. For analysis, Google Cluster Dataset is used. In order to improve the accuracy, data were normalized, and random samples of data were selected for clustering. The machine learning algorithms applied were able to successfully determine the appropriate clusters that can further be used for energy-efficient resource scheduling.

PRC6: Hybrid lightweight cipher for enhanced cloud data security in parallel environment

AbstractModern technologies of computing cloud are showing great promise, but at the same time create new security challenges that hinder full acceptance. Given that most of these services often use cloud networks as channels for communication, securing data transmission is crucial. This paper introduce a new hybrid encryption algorithm, the proposed two‐layered PRC6 cipher, tailored address security concerns in cloud computing environments with minimal resource constraints. The PRC6 cipher incorporates enhancements from Cha‐cha into an extension of the RC6 cipher. PRC6 implements double encryption. At the first level, the plain text is divided into four equal parts, each encrypted by processes derived from RC6, which include shifting, summation, modulo arithmetic, and XOR with a generated key. The second level incorporates a Quarter round function, among others, to further obscure the encoded message. PRC6 is implemented in a parallel computing model to significantly reduce overall computation time, especially important for lightweight applications. Experimental results show that the algorithm can achieve a high level of security for cloud workloads. It activates parallel mode in just seven encryption rounds, cutting calculation time to 50% in a matter of seconds. Performance evaluations against popular encryption standards also indicate that PRC6 offers promising security benefits when computational resources are limited. This hybrid approach presents a viable solution for strengthening data protection in modern cloud systems and it stand against the most popular attacks like brute force.

Design and Evaluation of a Hierarchical Characterization and Adaptive Prediction Model for Cloud Workloads

Design and Evaluation of a Hierarchical Characterization and Adaptive Prediction Model for Cloud Workloads

Load Balancing using Weight Based scheme in AWS

This paper presents research on load balancing methods for cloud computing platforms. Load balancing is an important element of cloud systems since it allows for optimal resource utilization and consistent performance levels. The study investigates several load balancing technologies and evaluates their effectiveness in coping with the dynamic and heterogeneous nature of cloud workloads. The study conducts experiments and analyses to discover significant parameters influencing load balancing performance, such as workload allocation, server capacity, and network latency. With the weight-based technique, different servers are dynamically given requests based on their weights, which are determined by the processing capacities of the servers. By distributing the work proportionately, this approach aims to decrease user request response times and avoid server overload. The effectiveness of the proposed approach is evaluated via simulations, demonstrating its potential to improve system performance and scalability relative to traditional load balancing methods. Generally speaking, this paper presents an optimized load distribution approach that raises the efficiency and stability of cloud computing systems. The findings enable to improve load balancing methods matched to the particular requirements and issues of cloud computing, eventually enhancing system scalability, reliability, and user

Cost-aware cloud workflow scheduling using DRL and simulated annealing

Cloud workloads are highly dynamic and complex, making task scheduling in cloud computing a challenging problem. While several scheduling algorithms have been proposed in recent years, they are mainly designed to handle batch tasks and not well-suited for real-time workloads. To address this issue, researchers have started exploring the use of Deep Reinforcement Learning (DRL). However, the existing models are limited in handling independent tasks and cannot process workflows, which are prevalent in cloud computing and consist of related subtasks. In this paper, we propose SA-DQN, a scheduling approach specifically designed for real-time cloud workflows. Our approach seamlessly integrates the Simulated Annealing (SA) algorithm and Deep Q-Network (DQN) algorithm. The SA algorithm is employed to determine an optimal execution order of subtasks in a cloud server, serving as a crucial feature of the task for the neural network to learn. We provide a detailed design of our approach and show that SA-DQN outperforms existing algorithms in terms of handling real-time cloud workflows through experimental results.

Cloud Workload Prediction Based on Bayesian-Optimized Autoformer

Cloud Workload Prediction Based on Bayesian-Optimized Autoformer

CarbonScaler: Leveraging Cloud Workload Elasticity for Optimizing Carbon-Efficiency

Cloud platforms are increasing their emphasis on sustainability and reducing their operational carbon footprint. A common approach for reducing carbon emissions is to exploit the temporal flexibility inherent to many cloud workloads by executing them in periods with the greenest energy and suspending them at other times. Since such suspend-resume approaches can incur long delays in job completion times, we present a new approach that exploits the elasticity of batch workloads in the cloud to optimize their carbon emissions. Our approach is based on the notion of "carbon scaling," similar to cloud autoscaling, where a job dynamically varies its server allocation based on fluctuations in the carbon cost of the grid's energy. We develop a greedy algorithm for minimizing a job's carbon emissions via carbon scaling that is based on the well-known problem of marginal resource allocation. We implement a CarbonScaler prototype in Kubernetes using its autoscaling capabilities and an analytic tool to guide the carbon-efficient deployment of batch applications in the cloud. We then evaluate CarbonScaler using real-world machine learning training and MPI jobs on a commercial cloud platform and show that it can yield i) 51% carbon savings over carbon-agnostic execution; ii) 37% over a state-of-the-art suspend-resume policy; and iii) 8 over the best static scaling policy.

Workload prediction for enhancing power efficiency of cloud data centers using optimized self‐attention‐based progressive generative adversarial network

SummaryNowadays, future workload prediction is an important requirement in cloud data centers to maintain flexibility and scalability of resources. However, due to unexpected peaks, drops in workload, noise, and redundancy in user requests, there is a considerable variance in resource demands, making it difficult to accurately predict workloads. Therefore, a self‐attention‐based progressive generative adversarial network (SAPGAN) optimized with Giza Pyramids Construction Algorithm (GPCA)‐based workload prediction is proposed for Sustainable Cloud Data Centers (CDC). At first, redundant data in historical data obtained through CDC are filtered utilizing Markov chain random field (MCRF) co‐simulation method. These pre‐processed historical data are supplied to the SAPGAN. The SAPGAN weight parameters are optimized by GPCA. The proposed method is analyzed using 2 benchmark datasets: HTTP traces from Saskatchewan and NASA. The simulation is implemented in JAVA. The performance metrics is examined to verify the efficacy of the proposed technique. The performance of the proposed approach provides 28.70%, 11.87%, and 14.79% higher accuracy; 30.15%, 11.72%, and 18.34% lesser energy consume for the dataset of NASA; and 5.32%, 2.45%, and 5.67% higher accuracy; 12.36%, 24.24%, and 34.16% lesser energy consume for the dataset of Saskatchewan HTTP traces compared with existing methods, such as auto‐adaptive learning in a dynamic cloud environment (AADEA‐WLP‐CDC), a neural network model depending on biphase adaptive learning for anticipating cloud data center workload (BALNN‐WLP‐CDC) and multiple scale ensemble of deep learning framework for multistep‐ahead cloud workload prediction (EMD‐LSTM‐GAN‐WLP‐CDC).

RAPID: Enabling fast online policy learning in dynamic public cloud environments

Resource sharing between multiple workloads has become a prominent practice among cloud service providers, motivated by demand for improved resource utilization and reduced cost of ownership. Effective resource sharing, however, remains an open challenge due to the adverse effects that resource contention can have on high-priority, user-facing workloads with strict Quality of Service (QoS) requirements. Although recent approaches have demonstrated promising results, those works remain largely impractical in public cloud environments since workloads are not known in advance and may only run for a brief period, thus prohibiting offline learning and significantly hindering online learning. In this paper, we propose RAPID, a novel framework for fast, fully-online resource allocation policy learning in highly dynamic operating environments. RAPID leverages lightweight QoS predictions, enabled by domain-knowledge-inspired techniques for sample efficiency and bias reduction, to decouple control from conventional feedback sources and guide policy learning at a rate orders of magnitude faster than prior work. Evaluation on a real-world server platform with representative cloud workloads confirms that RAPID can learn stable resource allocation policies in minutes, as compared with hours in prior state-of-the-art, while improving QoS by 9.0x and increasing best-effort workload performance by 19%–43%.