Search

Testing hypotheses about service quality management in IT infrastructure requires large and complex data centers with sufficient resources to explore various possible scenarios of infrastructure operation during the provisioning of IT services. For testing purposes, dozens of solutions already exist, but all of them don’t consider critical aspect of IT infrastructure. In order to solve this issue general mathematical model for quality of service management in critical infrastructures was introduced. Based on the proposed model simplest set of tools was developed for creating heavy simulations which can cover criticality during functioning.

IT-services are corner stones of any computing IT-infrastructures. In order to make them working properly their quality should be evaluated, measured and controlled. A lot of researches were conducted on quality of service evaluation, and many approaches were discovered. However, there is still need for connecting user satisfaction as subjective parameter with objective parameters or key performance indicators.Possible way for connecting those characteristics is to use marginal utility function in order to define direct dependency between quality of experience and quality of service. This kind of solution was proposed in this research.

The task of designing relational databases has always been the subject of scientific research, as it is associated with a number of interrelated steps. The result of each step is the development of models for presenting the future database with further refinement and, finally, the creation of an adequate relational database as a set of relations with the corresponding links between them. The article focuses on the normalization of databases, as one of the steps to create a datalogical model, namely, the use of the first three Normal Forms. As a result of the analysis carried out in the article, it was concluded that the definition of the Second Normal Form can be modernized and thus achieve two goals: to ensure the correct creation of potential primary keys and, thereafter, the correct external connections between relations, even before creating the data schema in a specific relational database. Moreover, to reconsider the necessity of applying the so-called “strengthened” or a higher version of the Third Normal Form, which speaks of mutual dependencies between key and non-key attributes.

Each IT service tends to solve its specific task with some predefined level of quality. In addition, providers of such services assure end users with some quality level according to the bucket user has bought. However, in order to provide those services at stated level it is important to know which kind of IT infrastructure does the provider use and which mathematical models can be applied to the hardware functioning, which is used for creating IT infrastructure. This article suggests using of artificial neural networks for classification and forecasting problems, which appear in IT infrastructure during provisioning of IT services. This can be made with indirect connection between IT resources usage and quality of service. Each IT service can have its own quality, which can be evaluated based on subjective and objective indicators of their performance. General problem, which can be solved in scope of the topic, is regression prediction problem, which can be perfectly solved with the use of neural networks. This paper presents neural network approach with decomposed groups of quality indicators, which implies breaking down IT infrastructure into hierarchical levels and defining quality indicators on each level with the further use of multilayer perceptron and recurrent neural network. Experimental results were compared with each other and have proven their effectiveness. The advantage of the use of neural networks in proposed problem is in small decline of predicted results from actual data.

The Internet of Things (IoT) is an emerging technology that offers great opportunities that is designed to improve the quality of consumers' lives, and also to improve economic indicators and productivity of enterprises, and more efficient use of resources. IoT system refers to the use of interconnected devices and distributed subsystems to leverage data gathered by sensors and actuators in some sort of environment and to take a proper decision on a high level. In this chapter, the authors propose an approach to Microcloud-based IoT infrastructure management to provide the desired quality of IT services with rational use of IT resources. Efficiency of IT infrastructure management can be estimated by the quality of services and the management costs. The task of operational service quality management is to maintain a given level of service quality with the use of minimum IT resources amount in IoT environment. Then, the maximum efficiency can be achieved by selecting such control when actual level of service corresponds to the coordinated with business unit and can be achieved by minimal costs. The proposed approach allows the efficient use of resources for IT services provision in IoT ecosystem through the implementation of service level coordination, resource planning and service level management processes in an integrated IT infrastructure management system based on hyperconvergence and software-defined principles. The main goals of this chapter are to investigate the state of art of the IoT applications resource demands in the context of datacenter architecture deployment and to propose Microcloud-based IoT infrastructure resource control method.

Modern applications in cloud computing, internet of things and machine learning are I/O intensive. They use data storage systems as the main resource of a data center. The advent of new storage technologies makes it possible to increase the performance of I/O operations by integrating devices with different performance within the data storage system by utilizing the storage-tiering approach. To prevent data loss and service downtime, the data storage systems must ensure fault tolerance using data replication management. As modern hybrid IT infrastructures are based on hyperconverged systems, the development of new methods and models for storage management in order to ensure high performance of I/O operations, high availability and fault tolerance becomes an urgent need. The authors propose the management method based on the model of a distributed two-level data storage system. The proposed method uses the algorithms for data migration between fast and slow levels of the data storage system and the algorithms for replication of data between the nodes of distributed data storage. The simulation results indicate that the proposed management method allows increasing performance of I/O operations with files and evenly placing replicas of data blocks on the data center nodes.

nowadays business cannot survive without integration with information technologies. That is why there is an emerging need in implementation and provisioning of IT services to end users and corporate users. However, with information technologies used business faces new challenges like appropriate and instant reaction to faults in process of provisioning of those services. That is why it is important to have automation tools for intelligent maintenance of service quality. This need is explained by the tight connection between resources required for providing services and their quality. In this paper, methods for quality evaluation and quality management were proposed. Quality evaluation according to proposed method works through the neural network classifier and management method uses fuzzy logic as a background framework. Experimental results have proved performance of the method.

Cloud data centers are most dynamic systems in a modern digital world. To deliver the high-performance and fault-tolerant IT services to end users effectively it is necessary to develop new methods for data center resource management while adapting to the emergence of new requirements. In this paper, the authors refine and evaluate the previously proposed method for cloud data center resource management based on the reinforcement learning approach. The proposed method takes into account the power consumption and the number of SLA violations in the management policy. The power consumption is managed by switching physical servers to active or sleep state depending on current utilization of three resources: CPU, memory, and network bandwidth. The proposed reinforcement learning agent allows to determine the optimal policy for managing the physical servers without creating an environment model and preliminary information about the workload. The evaluation results show that the proposed method allows to decrease the SLA violation time, to serve more VM schedule requests when the number of VMs is changing frequently, and to decrease the utilization of data center network due to decreased number of migrations.

The method for solving the problem of determining the influence of the faults in the access network on the service quality is presented. The solution of this problem is very important for large telecommunication service operators. The method combines rule-based approach and algorithmic approach to solve the problem in IP networks with RIP and OSPF protocols. The model of a network node, the algorithm and the rules for the considered routing protocols are developed. The evaluation of the proposed method is performed by modeling of large access network. The results show the operability and effectiveness of the proposed method.

Modern data centers’ providers are gradually moving away from traditional and multi-vendor IT infrastructures to open, standardized and interchangeable solutions that are based on a software defined approach to managing data center resources. The authors analyze the architectural features, requirements, limitations, hardware and software of hyperconverged infrastructures and their advantages in comparison with traditional and converged architectures deployed in data centers. The authors propose to employ two-level coordination schema to manage compute, storage, network and virtualization subsystems of hyperconverged infrastructure along with the self-management algorithms inside these subsystems.

A modern cloud data center is represented as a complex system where virtual machine consolidation and scheduling influence directly the cloud cost and performance. The virtual machine consolidation is the subject to many constraints originating from multiple domains, such as the resource requirements, user Service Level Agreement (SLA) compliance, security requirements, availability requirements, and other. Properly defined resource management methods and algorithms allow to achieve execution efficiency, SLA compliance, utilization of resources, energy saving, and the increasing profit of cloud providers. In this paper, the authors propose two versions of the Optimization using Simulated Annealing (OSA) algorithm to solve dynamic virtual machine consolidation problem. The virtual machine consolidation problem is considered as a multi-dimensional vector bin-packing problem. The authors take into account that the properties of items can be changed, new items may be requested to be deploy, and existing items may need to be reassigned to bins. Other constraints should be taken into consideration to solve virtual machine consolidation problem such as balanced load of resources of each physical machine, the limitation on maximum number of simultaneous migrations per physical machine, hardware constraints and other. The configuration of the system, the function for obtaining new configuration, the objective function for the optimization problem are determined for the proposed algorithms. The evaluation results show, that using OSA algorithms the simulated data center consumes almost the same amount of energy as while using a not optimized algorithm. On the other hand, the OSA algorithm with constraints allows to decrease overall performance degradation by virtual machines due to migrations, as a result, SLA violation is decreased. Furthermore, both OSA algorithms allow to reserve some resources of physical machine to react to increasing random resource demands in the nearest future.

The complexity of cloud data centers requires advanced resource management solutions that are capable to adapt to the changing customer's demands while providing continuous service and performance guarantees. The proposed Integrated Resource Management approach for energy-aware resource management provides the necessary elasticity on the physical machine level by considering a data center power consumption and a service-level agreement violation. The Integrated Resource Management approach is based on the data center state-space dynamic model, power consumption model, prediction model, SLA violation model, and capacity planning model method. The modeling process takes into consideration heterogenous environment, virtualized resource granularity, virtual machine live migration power consumption, virtual machines on/off state, and virtual machine deployment delay. The analysis and simulation, using Google cluster-usage traces, show that using Integrated Resource Management approach cloud service providers can achieve energy savings while minimizing SLA violations in terms of the number of overloaded physical machines and VM deployment delay.

Software-defined IT infrastructure does not have a clear definition in modern science world. There are many different interpretations among the corporate definitions of a software-defined IT-infrastructure, each of which is surely possible. General representation of software-defined IT-infrastructure was proposed based on those corporate interpretations. To solve the management problem in software-defined IT infrastructure the base management model was introduced. The base management model consists of three management loops namely outer, inner and operational loops. An analysis of several existing methods for service level management was conducted. Among the methods, which were considered in the paper, there are the use of management system with coordinator, injection of neural network into management subsystem, the use of a genetic algorithm, etc. The management system model and method for service level management, which is used in proposed management system, were developed basing on conducted overview of management methods and defined term of software-defined IT infrastructure. The method for management of software-defined IT infrastructure uses algorithm, which consists of three steps that can be executed separately or consequently. Those steps are reorganization of tasks, coordination of tasks according to IT infrastructure characteristics and adaptive planning, which is conducted based on actual data about IT infrastructure functioning. In order to study the efficiency of the proposed method the modelling with CloudSim was performed. Results of work were used in the management system of IT-infrastructure, which is named Smartbase ITS Control.

According to Forbes collected data, the public cloud services market will rise in 2016 and is expected to grow to $105 billions and, in fact, Virtual Machines (VMs) are one of the most popular tools for building “Cloud” services. But in service delivery such composition of multiple components greatly complicate analysis of product quality, especially if it is about Voice over IP (VoIP) set of services, which performance essentially depends on network quality. We provide a solution, that allows to perform deep network packet inspection with quality of service measurements with some quality of experience predictions, like some form of mean opinion score for VoIP. All system consists of a few completely open-source projects, developed with several professional teams. Currently system support Linux and QEMU/KVM-based solutions and is completely stream-based because of impossibility to save such amounts of data on 10Gbit link.

Virtual machine consolidation and scheduling influence directly the cloud cost and performance. They play an important role in cloud service granting helping to achieve execution efficiency, user Service Level Agreement (SLA) compliance, utilization of resources, energy saving, and the increasing profit of cloud providers. In this paper the authors propose the Optimization using Simulated Annealing (OSA) algorithm to solve dynamic virtual machine consolidation problem. The virtual machine consolidation problem is presented as an extension of the bin-packing problem. The configuration of the system, the function for obtaining new configuration, the objective function for the optimization problem are determined for the proposed simulated annealing algorithm. The evaluation results show that using OSA algorithm the simulated data center consumes almost the same amount of energy as not optimized algorithm, but OSA algorithm allows to decrease SLA violation and to reserve some resources of physical machine in order to react on increasing random demands in the nearest future.

The objective of this paper is to propose the rule-based algorithmic approach for solving problems of faults influence on the service quality in access networks. In this paper the graph oriented model for network and communication services representation, the rule-based system and corresponding output mechanism, the algorithms for the faults impact analysis based on graph traversal schemes are proposed. The proposed approach allows to obtain agility of rule-based systems and efficiency of arithmetic calculations. Essentially, the algorithm for estimation of faults impact preliminary is deduced with the help of rule-based system. The evaluation of the proposed approach is performed by modeling large access networks with a large amount of equipment and results show high performance of the algorithm.

A novel approach for service level management of corporate IT infrastructures is considered. Decomposition-compensation method of service level management of corporate IT infrastructures is proposed in this work. The method assumes the decomposition of tasks related to service level management and the compensation of negative impact of various factors by allocating extra resources for critical applications. The approach is based on the interaction of three integrated hierarchical processes—matching the level of services, resource planning, and service level management.

The Internet of Things (IoT) is an emerging technology that is designed to improve the quality of consumer's life and also to improve economic indicators and productivity of enterprises, for more efficient use of their resources. In this paper the authors propose an approach to Microcloud-based IoT infrastructure management which provides the required quality of IT services with rational use of IT resources. The proposed approach is based on decomposition-compensation method in which the task of operational service quality management is to maintain a given level of service quality with the use of minimum IT resources amount in IoT environment. Thus the maximum efficiency can be achieved by selecting such control when actual level of service corresponds to the agreed one with the business unit and can be achieved by minimal costs. The proposed approach allows the efficient use of resources for IT services provision in IoT ecosystem through the implementation of service level coordination, resource planning and service level management processes in an integrated IT infrastructure management system. The main goal of this paper is to propose Microcloud-based IoT infrastructure resource management approach using decomposition-compensation method.

The Internet of Things (IoT) is an emerging technology that is considered by industry and academy as a global infrastructure for the information society in today's digital world. IoT is implemented now in several ecosystems to improve economic indicators and productivity of enterprises, and also to improve the quality of consumer's lives. Thus a development of new architectures and approaches for IoT infrastructure management is needed. In this article the authors propose an approach to develop the Microcloud-based IoT infrastructure architecture based on decomposition-compensation approach [1] to provide the desired quality of IT services with rational use of IT resources. Efficiency of IT infrastructure management can be estimated by the quality of services and the management costs. The main goal of this article is to develop an architecture of management system for IoT infrastructures.

A widespread use of the cloud computing paradigm has increased the necessity and significance of improving the management efficiency of cloud infrastructures. Special attention is paid to solving cloud resource management problems. In this paper, authors present an architecture of Software Defined Cloud Infrastructure management system that leverages Software Defined approach in all subsystems: network, storage, and computation. Due to the intensive changes of virtual machine (VM) workloads and different conditions of resource utilization the VM placement and migration problems should be solved and optimized continuously in an online manner. To address such problems the authors propose novel heuristics for VM placement and consolidation based on a physical machine (PM) workload prediction. The authors also evaluate a particular policy of the VM allocation in a data center using an adaptive genetic algorithm. The proposed adaptive Software Defined approach to the cloud infrastructure management is implemented by using the policy selector that allows to select different algorithms or policies for resources and virtual machines management in order to adapt to the impact of disturbing influences.

A widespread use of the cloud computing paradigm has increased the necessity and significance of improving the management efficiency of cloud data centers. Special attention is paid to solving cloud resource management problems. Due to the intensive changes of virtual machine (VM) workloads and different conditions of resource utilization the VM placement and migration problems should be solved and optimized continuously in an online manner. To address such problems the authors present an approach to continuous new VM allocation and VM migration. The authors also evaluate a particular policy of the VM allocation in a data center using an adaptive genetic algorithm. The proposed Adaptive Software Defined approach to the cloud data centers management is implemented by using the policy selector, that allows to select different algorithms or policies for resources and virtual machines management in order to adapt to the impact of disturbing influences.

The decomposition-compensation method of service level management used in corporate IT infrastructures is proposed in the paper. The method assumes the decomposition of service level management tasks and the compensation of the negative impact of various factors by allocating extra resources for critical applications. The approach is based on the interaction of three integrated hierarchical processes — matching the level of services, resource planning and service level management. For the service level management the adaptive genetic algorithm was used.