Abstract: Load balancing is a key technology for supporting multi-tenancy SaaS applications requiring scalability. Given that current load balancing strategies are not suitable for SaaS, a load balancing strategy is proposed to support use of SaaS for service on demand applications. Based on analysis of customized user requirements, a set of rules and algorithms are proposed for system resource pre-allocation. Based on a system resource allocation protocol, a set of algorithms is proposed to provide limited customer request allocation. Experimental results showed that the load balancing strategy proposed herein satisfies the needs of users with different requirements and improves overall execution effectiveness.