An Autonomic Computing-based Architecture for Cloud Computing Elasticity

Elasticity is an important feature of cloud computing, and can be understood as how a computational cloud fits to variations in their workload by provisioning and deprovisioning resources. Autonomic Computing brings many concepts quite useful in the construction of elastic cloud computing solutions, such as control loops and thresholds-based rules. This paper proposes an elastic architecture for cloud computing based on concepts of Autonomic Computing. For its validation, we designed two experiments using microbenchmarks, applied in both private and hybrid clouds. Results shown cloud computing and Autonomic Computing may work well together in the elasticity provisioning.