Elastic Multi-tenant Business Process Based Service Pattern in Cloud Computing

Elasticity is an essential property of cloud computing. It helps service providers to efficiently exploit cloud resources and reduce servicing costs. Therefore, the multitenant business processes are long-running and they are concurrently accessed by dynamic requests from tenants. However, ensuring business process elasticity at the infrastructure and the platform levels may only result in significant resources waste due to the partner services autonomy and variability. For this purpose, we tackle the problem of handling elasticity at the process and services levels to scale-out and scale-in their service instances whenever possible. To do this, we propose an auto-scaling approach to hold the promise of ensuring the elasticity of the multitenant business process. This research is based on service patterns for integrating multi-tenancy architecture and making ecisions on the execution of the elasticity mechanisms. Furthermore, we encapsulate our approach into a middleware layer (Middleware as a Service) between the application layer and the platform layer in the cloud architecture. Provided experimental evaluations show that our approach is efficient for ensuring elasticity under various workloads variation of the business process in cloud computing.