Proactive Learning from SLA Violation in Cloud Service based Application

Abstract

In recent years, business process management and Service-based applications have been an active area of research from both the academic and industrial communities. The emergence of revolutionary ICT technologies such as Internet-of-Things (IoT) and cloud computing has led to a paradigm shift that opens new opportunities for consumers, businesses, cities and governments; however, this significantly increases the complexity of such systems and in particular the engineering of Cloud Service-Based Application (CSBA). A crucial dimension in industrial practice is the non-functional service aspects, which are related to Quality-of-Service (QoS) aspects. Service Level Agreements (SLAs) define quantitative QoS objectivesandis a part of a contract between the service provider and the service consumer. Although significant work exists on how SLA may be specified, monitored and enforced, few efforts have considered the problem of SLA monitoring in the context of Cloud Service-Based Application (CSBA), which caters for tailoring of services using a mixture of Software-as-a-Service (SaaS), Platform-as-a-Service (PaaS) and Infrastructure-as-a-Service (IaaS) solutions. With a preventive focus, the main contribution of this paper is a novel learning and prediction approach for SLA violations, which generates models that are capable of proactively predicting upcoming SLAs violations, and suggesting recovery actions to react to such SLA violations before their occurrence. A prototype has been developed as a Proof-Of-Concept (POC) to ascertain the feasibility and applicability of the proposed approach.

Development of the method of features learning and training decision rules for the prediction of violation of service level agreement in a cloud-based environment

“…However, solution of this problem is complicated by a priori uncertainty regarding the functional state of the service as a result of non-stationarity of demand and heterogeneity of physical and virtual components of IT-infrastructure. Papers [3,4] suggested application of the algorithms of a decision tree, random forests and Naïve Bayes to remove uncertainty concerning compliance with SLA conditions, associated with exceeding service response time, service availability or a decrease in information safety. However, a small number of features were controlled in proposed approaches, which prevented obtaining a highly reliable predictive model for advances period of time sufficient for the use of necessary measures.…”

“…However, a small number of features were controlled in proposed approaches, which prevented obtaining a highly reliable predictive model for advances period of time sufficient for the use of necessary measures. The authors of [4,5] proposed to examine the trends of using resources within a sliding window of the assigned size for the formation of feature description of predicted functional states. In study [5], the prediction model is based on recurrent neural network of Long Short-Term Memory, LSTM.…”