Constraint-Based Multi-Tenant SaaS Deployment Using Feature Modeling and XML Filtering Techniques

Software-as-a-service (SaaS) is becoming more important as a software delivery and service model. However, multi-tenancy, which promises to provide a high degree of resource sharing among a large number of tenants (customers or organizations), can significantly complicate SaaS development, deployment, and management due to potential explosion of co-existing tenant-specific variations. Manually configuring those tenant-specific variations for multi-tenant SaaS systems cannot satisfy scalability and flexibility. In this paper, we propose a novel approach in support of multi-tenant SaaS systems using feature modeling and XML filtering techniques. Feature modeling is used to capture functional, non-functional requirements and constraints. The features of a cloud system and tenant-specific requirements are encoded with XPath feature representations and XML document, respectively. We adopt Yfilter, an established XML filtering technique, and tailor it to match cloud configurations that satisfy tenant-specific requirements and constraints. The experimental results demonstrate that our approach can automatically and correctly identify cloud system configurations that match tenant-specific requirements. In addition, the execution time in our approach is only a small fraction compared to the existing approaches (e.g., Fama) and the configuration space is also smaller.