Adaptive QoS Management Mechanism Of IT System

In recent years, IT plays an important role in modern enterprise, and the theory, method and practice of IT management need to be further devloped corespondingly. The QoS management issue is well known as one of the most important aspects in IT management research area. Owing to the increasing scale of IT systems and the complex and flexible implementation techniques in modern enterprise, an adaptive and intelligence method of the QoS management of IT system is needed. Moreover, with the relationship between IT and business of companies or organizations becoming more and more tightly, QoS management of IT system should consider business factors, i.e. business driven IT management (BDIM) method, so that IT can bring more business values to the enterprises or organizations.This paper focused on the QoS management issues in IT operation stage. It built an adaptive QoS management model and maintenance mechanism with the intelligent control theory in the traditional SLA description based enterprise IT environment. It also adopted the BDIM methodology to establish the mapping relationship between the business layer and IT layer from the business perspective and to implement the adaptive QoS management mechanism with the goal of business result. It can continuously provide the best business result. The main contributions of this paper are as follows:(1) For self-adaptive QoS (SLO) Management issues in IT systems with SLA description, this paper presents a self-adaptive QoS management model for IT system with closed-loop control theory. Based on the model which can enable dynamical regulation of the resources and capacity of IT infrastructure, this paper built a fuzzy feedback control method based self-adaptive regulation mechanism to maintain the QoS (SLO) stipulated in SLA on the operation stage of IT system.(2) Base on the adaptive QOS regulation mechanism of IT systems, this paper adopted the BDIM method to considerate the adaptive QoS optimization issues in IT systems scenario, and proposed a BDIM based adaptive QoS optimization mechanism. This mechanism can adaptively adjust the resources of IT infrastructure according to a set of well-defined business impact weight in order to obtain the maximal overall business utility, which enables IT continuously deliver high level values to enterprises and organizations.(3) Based on our proposed adaptive QoS optimization mechanism, this paper further focused on the adaptive QoS optimization issues in complex multi-level and multi-tire IT systems. This paper proposed an adaptive QoS optimization mechanism in the IT infrastructure virtualized environment, which allocated appropriate IT resources to the applications on each layer in a more efficient way on the operation stage of IT system. The mechanism not only made the QoS parameters satify the SLA requirements but aslo obtained the maximal overall business utility.(4) In the implementation process of adaptive QoS management mechanisms, this paper proposed a dynamic task decomposition algorithm based on multi-agent (MA) collaboration considering the requirements of adaptive QoS management mechanism and the characteristics of the management task itself. The proposed algorithm considered the capacity constraints of MA, and it can dynamically adapt the variation of QoS management task due to the changes of management policy, IT infrastructure or business logic. In addition, the algorithm ensures the balance of the decomposed sub-tasks, namely the work load balance of the MA. As result, it can improve the efficiency and stability of MA and effectively support the implementation of adaptive QoS optimization mechanism.