Service-oriented architecture (SOA) paradigm for the purpose of large-scale applications offers meaningful cost savings by rework existing services. However, the high oddity of client appeal and the allocated character of the access may depreciate service response time and chance. Static cloning of components in database for placing load spikes need efficient resource planning and also uses the cloud infrastructure. Moreover, no service chance gives trust is provided in situations like datacenter crashes. In this paper, we explore a cost-efficient usage for dynamic and geographically-diverse cloning of elements in a cloud computing infrastructure that perfectly adapts to load differences and provides service chance guarantees. When comes to economic level, components hire server opportunities and clones or trashes themselves based to self optimizing situations. We proved in real time applications that such an access better in response time even full cloning of the components in all servers, while providing service chance guarantees under failures.