Achieving Market Liquidity Through Autonomic Cloud Market Management

Abstract

Due to the large variety in computing resources, Cloud markets often suffer from a low probability of finding matches between consumers’ bids and providers’ asks, resulting in low market liquidity. The approach of service level agreement (SLA) templates (i.e., templates for electronic contracts) is a mean to reduce this variety as it channels the demand and supply. However, until now, the SLA templates used were static, not able to reflect changes in users’ requirements. To address this shortcoming, we introduce an adaptive approach for automatically deriving public SLA templates based on the requirements of market participants. To achieve this goal, we utilize clustering algorithms for grouping similar requirements and learning methods for adapting the public SLA templates to observed changes of market conditions. To assess the benefits of the approach, we conduct a simulation-based evaluation and formalize a utility and cost model. Our results show that the use of clustering algorithms and learning algorithms improves the performance of the adaptive SLA template approach.