Cloud Resource Demand Prediction using Machine Learning in the Context of QoS Parameters

Abstract

Predicting demand for computing resources in any system is a vital task since it allows the optimized management of resources. To some degree, cloud computing reduces the urgency of accurate prediction as resources can be scaled on demand, which may, however, result in excessive costs. Numerous methods of optimizing cloud computing resources have been proposed, but such optimization commonly degrades system responsiveness which results in quality of service deterioration. This paper presents a novel approach, using anomaly detection and machine learning to achieve cost-optimized and QoS-constrained cloud resource configuration. The utilization of these techniques enables our solution to adapt to different system characteristics and different QoS constraints. Our solution was evaluated using a system located in Microsoft’s Azure cloud environment, and its efficiency in other providers’ computing clouds was estimated as well. Experiment results demonstrate a cost reduction ranging from 51% to 85% (for PaaS/IaaS) over the tested period.

CLR: Cloud Linear Regression Environment as a More Effective Resource-Task Scheduling Environment (State-of-the-Art)

“…CLR model going to be implemented on all cloud-VMs and taking into account its logical dependence on building an integrated machine learning model that complies with the requirements of cloud computing. Hence, CLR logically includes the following [18,19,20,21]:…”

“…It is used for the examination of efficiency and performance in a real-world cloud. Logically, any update or change in the dataset-behavior or dataset-nature is reflected in the performance of scheduling and resource-allocation policies [20]. Due to the type of users' data confidentiality and policies in the open cloud environment, the real cloud workload is hard to measure for performance analysis.…”

“…Currently, many feature selection algorithms have been deployed such as evolutionary algorithms (e.g. Particle Swarms Optimization, Ant Colony Optimization) and stochastic approaches [20,21]. Some of them are based on classical methods like simulated annealing and genetic algorithm.…”

Anomaly detection in the context of long-term cloud resource usage planning

“…Our approach is similar in terms of data preprocessing; however, our anomaly detection mechanism does not filter out anomalies which occur with a certain constant frequency, as this indicates that these changes are in fact not anomalous and should be taken into account when predicting future usage. The authors of [20] present an approach that uses anomaly detection and machine learning to achieve cost-optimized and QoS-constrained cloud resource configuration. The solution developed was tested using a system located in Microsoft's Azure cloud environment.…”

Virtual Machine Instance’s Price Prediction using Machinelearning Techniques at the Cloud Data Center

“…First, they formulated the problem using Integer Linear Programming (ILP) technique, then they allocated and migrated the VMs using a cut and solve based algorithm and call back method. Nawrocki et al [11] proposed an anomaly detection and machine learning based resource prediction technique at the cloud data center. In their proposed approach by predicting the resource requirement they optimized the cost and also achieved the QoS at the cloud data center.…”