Models to evaluate service provisioning over cloud computing environments-a blockchain-as-a-service case study

Abstract

The strictness of the Service Level Agreements (SLAs) is mainly due to a set of constraints related to performance and dependability attributes, such as availability. This paper shows that system’s availability values may be improved by deploying services over a private environment, which may obtain better availability values with improved management, security, and control. However, how much a company needs to afford to keep this improved availability? As an additional activity, this paper compares the obtained availability values with the infrastructure deployment expenses and establishes a cost × benefit relationship. As for the system’s evaluation technique, we choose modeling; while for the service used to demonstrate the models’ feasibility, the blockchain-as-a-service was the selected one. This paper proposes and evaluate four different infrastructures hosting blockchains: (i) baseline; (ii) double redundant; (iii) triple redundant, and (iv) hyper-converged. The obtained results pointed out that the hyper-converged architecture had an advantage over a full triple redundant environment regarding availability and deployment cost.