A Framework for SLA Violation Prevention in Cloud of Things Environment

The existing approaches use runtime monitoring to detect Service Level Agreement (SLA) violation and prediction techniques for SLA violation prediction by using the historic QoS data to predict the future QoS values. However, these approaches do not consider the occurrence of events and the impact they may have on SLA violation. Moreover, existing approaches also do not provide any recommendations to the SLA manager to prevent violation proactively. These limitations of the current literature require the development of a comprehensive framework that can identify and capture events which impact a services quality/performance and model their effect on QoS attributes to reduce the negative consequences by avoiding SLA violation. To address these problems, we propose an SLA violation prevention framework. The proposed framework provides a methodology for recommending suitable actions to the SLA manager for SLA violation prevention. Moreover, the performance and the usefulness of the proposed framework is evaluated and validated by developing a proof of concept and applying it on case studies.