Simulation of Techniques to Improve the Utilization of Cloud Elasticity in Workload-aware Adaptive Software

Abstract

More and more software owners consider moving their IT infrastructure to the cloud. At present, cloud providers offer easy manners to deploy software artifacts. Therefore, the profile of cloud clients is no longer limited to computing experts. However, an appropriate configuration of the elasticity offered by cloud computing is still complicated. To help these clients, this work presents a simulator of the behavior of software services that run on the cloud and use the cloud elasticity for adapting their infrastructure in order to accommodate their workload in each moment. This work identifies techniques that are used to help mitigating at runtime the lack of predictability of workload changes. The presented simulator implements the identified techniques and allows users to execute scenarios where a combination of these techniques is enabled.